(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "The foundry sands of Minnesota"

hGGtob 




©fj? Itttu?r0ttg nf iltmteaoia 



MINNESOTA GEOLOGICAL SURVEY 
William H. Emmons, Director 

BULLETIN NO. 1 18 



THE 
FOUNDRY SANDS OF MINNESOTA 



BY 



G. N. KNAPP 





J W 






/' 



MINNEAPOLIS 

The University of Minnesota 

1923 









)S J frfio 



10. ) 8 



BOOK 557.76.M666B no. 18 c. 1 
MINNESOTA GEOLOGICAL SURVEY # B 
ULLETIN 



3 T153 0D13bTMfl T 



Digitized by the Internet Archive 
in 2013 



http://archive.org/details/foundrysandsofmiOOknap 



Sty? Jtnttier0tig of JMimtfanta 

MINNESOTA GEOLOGICAL SURVEY 
William H. Emmons, Director 

BULLETIN NO. 18 



THE 
FOUNDRY SANDS OF MINNESOTA 



BY 



G. N. KNAPP 




MINNEAPOLIS 

The University of Minnesota 
1923 



''•'" 1a 



PREFACE 

This paper, on the sands of Minnesota, is a report of an investigation 
begun in 191 8, at a time when there was a great demand for information 
about local molding sands, as a result of the traffic situation that made it 
difficult or impossible to obtain sands from sources that had previously 
supplied them. The inquiry, which was in charge of Mr. G. N. Knapp, 
showed that Minnesota contains an abundance of sands for founding nearly 
all products, equal to, or better than, the sands that had been imported. 
Part of the information embodied in the report was placed at the disposal 
of the metal founders as soon as it became available, either verbally or by 
means of mimeographed sheets. 

The field work showed also that molding materials are more widely 
spread over the state than was supposed, and a fairly comprehensive 
reconnaissance was made of a considerable part of Minnesota. Materials 
gathered from widely separated sources were tested in the laboratory and 
some were tested by founders in their plants. Minnesota contains an 
abundance of foundry sands and all materials commonly used for making 
molds in foundries of iron, steel, brass; and aluminum, except possibly a 
highly plastic refractory clay. 

An economic bulletin such as this is chiefly utilitarian. It should show 
the nature of materials used, the tests made to ascertain the value of the 
materials, and the distribution of the materials in the state. These subjects 
are taken up separately in this bulletin and the geologic and geographic 
distribution are discussed in some detail. The report does not contain a 
geologic map. To have included such a map would have resulted in dupli- 
cation. A search for sands in the hard rock formations will be aided by 
the use, in conjunction with this bulletin, of the county geologic maps, 
published by N. H. Winchell and associates in the final report, Volumes I 
to V, Geological and Natural History Survey of Minnesota.- The search 
for materials in the drift will be aided by the maps of surface formations 
and agricultural conditions, by F. Leverett and F. W. Sardeson, published 
by this survey, and included in Bulletins 12, 13, and 14, already issued. 
Bulletin 11, on the clays and shales of Minnesota, by Professor F. F. 
Grout and Dr. E. K. Soper, 1 will be found helpful in a search for clay 
binders, and Bulletin 663, of the United States Geological Survey, by 
Mr. Oliver Bowles, will show the sources of screenings from quarry 
operations, such as are used to a limited extent. 

1 Reissued as Bulletin 678, United States Geological Survey. 



iv PREFACE 

The economic bulletins of this survey are written principally for the 
use of the general public. We aim to state the results of investiga- 
tions in terms that are non-technical. If we fail in this respect, and the 
results of the investigation are obscure, or if they do not contain the in- 
formation desired, the officers of the survey are available for consultation. 
The survey charges no fee for such service. 

William H. Emmons. 



CONTENTS 

Pages 

Preface iii-iv 

Introduction 1-3 

Acknowledgments 3 

Founding 3-6 

Nature of sands and loams for foundry molding 6-8 

"Doctoring" molding sands. 8-9 

Outline of geology of Minnesota 9-17 

Archean period 13 

Algonkian period 13 

The Red Clastic series 13 

Cambrian period 14 

Ordovician period H -1 ^ 

Devonian period 15 

Cretaceous period 15-16 

Pleistocene 16-17 

Recent 17 

Geological formations yielding foundry sands, loams and clays. . . 17-30 

The Sioux quartzite formation 17-18 

The Kettle River formation 18-19 

The Jordan formation 19-21 

The St. Peter formation 21-22 

The Decorah formation 22 

The Cretaceous formation 22-25 

Glacial deposits . . 25 

Moraine deposits 25-26 

Glacial loams . 26 

Outwash deposits 26-27 

Lake beaches 28 

Loess 28-29 

Recent deposits ^o 

List of Minnesota foundries and localities supplying them with 

foundry sands 3 I_ 3 2 

Laboratory methods of testing foundry sands, loams, and clays 32-48 

Mechanical analyses of sands, loams, and clays 32-36 

Comparison of wet and dry methods of analysis 36-37 

Comparing analyses 37 _ 39 



CONTENTS vi 

Pages 

Laboratory methods of testing foundry sands, loams, and clays 
(continued) 

Laboratory tests for bonding power of sands, loams, and clays 39-48 

Determining porosity of foundry sands and loams 4 I_ 43 

Testing the permeability of molding sands 43-48 

Apparatus 44 _ 46 

Mounting samples 46-48 

Relation of porosity, texture, and structure to permeability 48-53 

Effective size of grain 53 _ 54 

Method of computing percentage permeability 54-55 

Shape and arrangement of interstitial tubes in sand 56-57 

Cores used in permeability tests of sands, loams and clays 56-57 

Fusion tests 57 

Moisture content of molding sands 57-58 

Mechanical and mineral analyses of core sands 58-62 

Mechanical and mineral analyses of steel sands 62-66 

Mechanical and mineral analyses of brass sands 66-69 

Mechanical and mineral analyses of loams 69-75 

Synthetic sands and loams 72 

Clays 75-82 

Colloidal material in clays 76 

Plasticity of clays 7^~77 

Tensile strength of clays yy 

Mechanical analyses of clays 77~82 

Laboratory tests of tensile strength 82-84 

Deterioration of sand with heat 84-85 

Tensile strength of molding sand briquets 86-88 

Permeability tests 89-94 

Results of permeability tests 95 _ 98 

Porosity tests 98-100 

Conclusions 100-101 



LIST OF ILLUSTRATIONS 



Pages 



Fig. i. Section of Cretaceous clay at South Bend, 

Minnesota 23 

2. Map of southeastern Minnesota, showing area 

covered with loess 28 

3. Hot air engine used for shaking bottles in clay 

determination 34 

4. Brass mold for molding sand briquets for tensile 

strength tests 39 

5. Apparatus for breaking briquets of molding 

sand to determine brass tensile strength 39 

6. Cylinders used for mounting cores of sand for 

permeability tests, and cup used in porosity 
determinations 42 

7. Apparatus used in permeability tests delivering 

74 liters of air under constant pressure of 

50 inches of water 44 

8. Diagram showing the closest system of packing 

of uniform spheres that is possible 49 

9. A and B. Illustrating structure in mass of spheres in most 

open arrangement 5° 

10. A and B. Illustrating structure in mass of spheres in most 

compact arrangement 50 

11. Illustration of rhombohedron cut from mass 

of spheres in most compact arrangement, 
showing arrangement, relative size, and com- 
munication of interstitial spaces 51 

12. Illustration of granulation in which spherical 

grains are bound together in compact grains 
or granules arranged in open structure with 

large intergranular spaces 52 

13- Diagram showing relations of meshes of screens 

and sizes of openings 54 



THE FOUNDRY SANDS OF MINNESOTA 

INTRODUCTION 

When the Federal government in 191 7, as a war measure, assumed 
control of. the railroads of the nation, the founders of Minnesota were 
notified that the Railroad Administration would not continue to furnish 
cars for the shipment of sand from remote points, and were advised to 
find sand nearer home. The writer was detailed to investigate the prob- 
lem of where foundry sands were to be obtained in Minnesota. 

In the course of this. work, begun in June, 1918, the southern part of 
the state from Taylors Falls to the Iowa line and west to Pipestone and 
Granite Falls was covered in a reconnaissance way by the writer ; and 
Mr. V. T. Allen extended the work to foundries in the northern counties. 
Sands were found in various parts of the state similar in character and in 
some respects superior in quality to those which were then in use in the 
foundries and which had been imported from Missouri, Illinois, Wisconsin, 
Kentucky, and New York. 

Circulars were issued by the Minnesota Survey in 191 8 and 191 9 on 
brass sand, quartz sand, foundry loams, and refractory clays of Minnesota, 
and distributed to the foundries desiring such information. 

At the time this work was undertaken there were thirty-five foundries 
in Minneapolis and St. Paul actively engaged in casting the various metals. 
Some of them confined their operations exclusively to one line of work 
such as grey iron founding, steel founding, brass founding, or ornamental 
bronze founding; others covered a larger field, founding in grey iron, 
steel, semi-steel, brass, aluminum, and bronze. 

Some of the foundries were content to use such sands and loams for 
molding as could be found in the immediate vicinity of their shops, and 
could be hauled by team or truck. Others having more exacting contracts 
calling for high finish to their castings resorted to artificial mixtures of 
sands and loams, or sands and clays, partly imported and partly obtained 
locally. Still others had imported all of their sands and clays from 
remote regions. One foundry doing ornamental bronze work imported 
a facing sand from France at a cost of $16 per barrel. The sand imported 
from New York chiefly for brass work, known as "Albany sand," cost 
f.o.b. Minneapolis about $15 per ton, while sand equally good for the pur- 
pose was found in Minnesota, costing 25 cents per ton in the pit, or $3 to 
$4 per ton at St. Paul. 

Time did not permit gathering complete statistics of the amount of 
sand, loam, and clay used by the foundries of Minnesota, or the cost of 
the same, but from such statistics as were gathered it is roughly estimated 



2 THE FOUNDRY SANDS OF MINNESOTA 

that the total amount used annually exceeded 100,000 tons; and as the 
price to the foundries ranged from $2 to $15 per ton, it is apparent that 
the cost is a very considerable sum. 

The greatly increased demand within the past few years for a higher 
grade of casting, both as to the soundness of the casting and as to the 
finish or exterior surface, more especially in steel, brass, and malleable 
iron, has led to a great expansion of business in these lines. This demand 
was augmented by the contracts for war materials, and the rigid inspec- 
tion and exacting conditions insisted on by the government. 

Demand for a higher grade product has brought home to the founder 
the importance of the sand ; the grade of the cast product depends largely 
on the quality of the sand forming the mold. The percentage of defective 
castings in some of the foundries visited was so large that the loss incident 
to scrapping them came near making a loss on the total operation. The 
defects in the castings, in a large majority of cases at least, were plainly 
due to the quality of the sand used in the mold or to improper handling of 
the sand. In only a small percentage of cases were the defective castings 
due to the metal, itself, or to the methods of handling the metal. 

Most of the foundries were without laboratory facilities for testing 
their molding sands, except by screen analysis and chemical tests ; accord- 
ingly when confronted by sand trouble that did not yield to empirical 
methods, they tried another sand usually a new one sought from the 
locality where some competitor making better castings was obtaining his 
sand. 

In order to appreciate the problems connected with metal founding in 
Minnesota, most of the plants were visited and samples of the molding 
sands collected. These were studied in the laboratory. Later various 
geological formations within the state were studied in the field and samples 
collected from them. These were submitted to similar laboratory tests, 
and some of the most promising were supplied to foundries to be tested 
under actual working conditions. 

The laboratory tests to which the samples of sand, loam, and clay col- 
lected in the foundry and in the field were subjected consisted in (a) a me- 
chanical analysis, including a determination of the clay content by elutria- 
tion; (b) a test of the bonding power or tensile strength by breaking 
briquets made from the various materials; (c) determination of the 
porosity by comparing the weight of a given volume of sand with the same 
volume of non-porous rock; (d) a determination of the permeability by 
passing a measured volume of air under constant pressure through a core 
of sand of standard size; (e) a refractory test, which consisted in deter- 
mining the point of incipient fusion of the materials in an electrical fur- 
nace, the temperature of which was controlled; and (f) a mineral analysis. 



ACKNOWLEDGMENT 3 

The following report in which the results of the investigation are re- 
corded is divided as follows : the technology of the sand-consuming in- 
dustries in brief ; the geology of Minnesota with reference to the origin 
and distribution of sands, as well as some loams and clays ; a description 
of the laboratory methods and the apparatus used in testing; and a tabula- 
tion and discussion of the results of the laboratory tests. 

ACKNOWLEDGMENT 

The writer is indebted to Professor W. H. Hunter and Professor 
R. A. Gortner for suggestions in problems of chemistry ; to Professor 
H. A. Erikson for assistance in establishing tests for permeability and 
capillarity; to Professor F. F. Grout for aid in microscopic examination 
of silts, suggestions of laboratory methods, and criticism of the text ; 
to Mr. V. T. Allen for mineral analyses of the sands and notes on the 
results and data from foundries in the northern part of the state; to 
Mrs. W. C. Knapp for mathematical computations and general editorial 
work; to Mr. H. K. Armstrong for assistance in the laboratory; to 
Dr. W. A. Schaper for suggestions as to arrangement of material ; and 
to Professor W. H. Emmons for general direction of the work and for 
editing the report. 

Special acknowledgment is due managers and superintendents of vari- 
ous foundries in Minneapolis and St. Paul for courtesies shown ; to mem- 
bers of the technical force of these foundries for molding sand specifica- 
tions, chemical analyses of sands and loams, and other valuable data ; and 
to the skilled molders for the making of special cores and cylinders of 
molding sand and core sand for permeability tests. 

FOUNDING 

That certain metals were ductile, and by hammering could be fashioned 
into implements of war or the chase or into tokens and emblems, was dis- 
covered by primitive man in the early dawn of civilization, as evidenced 
by the relics found in the mounds and burial places of the ancient races. 

That the native metals melt at moderate temperatures and that by 
mixing them in varying proportions, alloys could be made having properties 
superior for certain purposes to the metals themselves, was no doubt dis- 
covered also by early man. This achievement was a prominent one in the 
"Bronze Age." 

The more important discovery that the ores of iron could be smelted 
and the metallic iron recovered in this manner, marked a still further ad- 
vance in civilization and was possibly made in prehistoric times, at least 
such evidence as we have indicates that it was made more than six thou- 
sand years ago. The oldest iron implements come from Egypt and date 



4 THE FOUNDRY SANDS OF MINNESOTA 

back more than live thousand years. The earliest recorded history has 
it that Tubal-Cain (B.C. 3875) was known as "master in every kind of 
bronze and iron work," indicating that the use of the metals had then 
reached an advanced stage. 

Founding, or the art of casting metals, is such a simple matter, now 
that we are familiar with the procedure, that it seems strange that the 
process was not discovered and taken advantage of as soon as metallic 
iron was recovered by smelting the ores. One would suppose that the 
next step after having discovered the process of smelting would have been 
to bring the metals into a fluid state and pour it into molds or forms, in 
this manner casting implements of any form or shape desired. 

It appears, however, that the first iron implements were made of 
wrought iron instead of cast iron ; that is, instead of melting the iron and 
pouring it into molds, it was taken from the furnaces in the form of billets 
of white-hot metal and hammered into the form desired. So far as now 
known, wrought iron was used for one thousand years, and possibly two 
thousand years, before the art of casting the metal was discovered, or at 
least before cast iron was used industrially. 

A discussion of the technique of molding and the methods of handling 
the melted metal has no place in the present report, except as it may aid 
those unfamiliar with the procedure better to appreciate the kind of ma- 
terial required for molding sand and for other foundry uses. A very brief 
outline of one of the simplest methods will suffice, and the reader, inter- 
ested in further details, may find them in one of the numerous books on 
foundry practice. 

If it is desired to make a sphere or ball of cast iron, or of other metal, 
the procedure is as follows : First a pattern is made, which is a counterpart 
of the ball to be cast. The pattern is commonly made of wood but may be 
made of other materials, such as clay, wax, etc. When made of wood the 
pattern is made in two parts, with dowel pins or other devices to hold 
them together and in alignment. 

After the pattern is made the second step is to make a mold from this 
pattern. One half of the pattern is placed on the foundry floor, molding 
bench, or table, with the flat face downward. Around the pattern is placed 
a frame of either wood or iron of convenient size. This is simply a box 
without cover or top into which is packed molding sand until the frame is 
level full. The molding sand is simply damp sand having enough clay or 
other binding material in it to hold it together after it is packed or rammed 
in firmly. The frame, filled with molding sand in which the half pattern 
is embedded, is next picked up carefully and turned over, bringing the flat 
face of the pattern uppermost. 

1 Moldenke, R., Principles of iron founding, p. 2. McGraw-Hill Book Co., New York, 1917. 



FOUNDING 5 

The other half of the pattern is then placed in position on the half 
buried in the sand. A second frame is placed upon the first ; the surface of 
the sand in the first is sprinkled with dry sand or dust to make a parting, so 
that the molding sand to be rammed in will not stick to the sand in the 
first frame. The second frame is then rammed full of molding sand level 
with the top. The upper frame, filled with sand, is then lifted off, and the 
halves of the pattern removed from the sand, leaving an impression, or 
"mold," in the sand the exact size and shape of the halves of the pattern. 
The two frames are again placed one upon the other in exact alignment so 
that the spherical cavity, or mold, in the center of the sand mass will be a 
sphere. The two frames are then locked together and weights are placed 
upon the upper surface of the sand. 

In ramming in the sand a small hole, called a "riser," is left from the 
face of the pattern to the surface of the sand in one of the frames. This 
serves as a passage through which the melted iron or other metal may be 
poured into the mold. 

The frames above described constitute a "flask." A flask is simply a 
rigid container made of two or more parts and serves to hold the rammed 
sand in position so that it may be lifted, carried around, or turned over. 
It also serves to reinforce the sand against the pressure of the metal and 
against the steam and gases generated in the sand when the hot melted 
metal is poured in. Flasks vary from simple one-piece boxes to complicated 
many-piece devices and in weight from a fraction of a pound to many tons. 

After the different parts of the flask are in position so that the mold 
is complete the melted metal is poured into the mold and allowed to stand 
for a period varying with the size of the casting from a few minutes to 
an hour or more. As soon as the metal is set, and usually while it is still 
red hot, the flask is lifted apart and the metal casting is shaken out on the 
foundry floor where it lies in the loose sand until cooled sufficiently to 
handle. 

As soon as the molding sand has cooled to room temperature, it is used 
again, the process being repeated as long as the sand has binder enough to 
stand up in the mold. Clay or new loam is added from time to time to 
replenish the bonding material which is destroyed by the heat. 

Some cast articles are so small or have so many reentrant angles, that 
it is not feasible to make two-piece patterns. In such cases a pattern is 
made of wax, which is a counterpart of the article to be cast. This wax 
pattern is embedded in the molding sand in a metal flask. The molding 
sand is then heated to a temperature sufficient to burn up or vaporize the 
wax which disappears leaving a mold in the interior of the sand mass, 
which in turn is filled by pouring the metal into the mold. This process 
is commonly employed in casting ornamental tablets and figures in bronze, 



6 THE FOUNDRY SANDS OF MINNESOTA 

aluminum, etc., and is extensively used in modern dentistry in casting gold 
fillings, and other restorations for teeth. In fact the adaptation of casting 
in metals, and the diversity of patterns, materials, and processes are almost 
unlimited. 

NATURE OF SANDS AND LOAMS FOR FOUNDRY MOLDING 

A great variety of earthy materials is used in the modern foundry 
under the general names of loam, clay, and sand. The loams used include 
material ranging from very fine silts to coarse sandy materials with a 
greater or less amount of gravel, and a clay content ranging from 2 to 10 
per cent. The sands used range in mineral content from pure quartz 
sand to material bordering on loams, in which a variety of materials is 
present, with more or less clay ; and in size of grain they range from 
material the bulk of which would just pass the 20-mesh screen, to material 
the bulk of which would pass the 100-mesh screen. Foundry clays range 
in physical constitution from the purest clay obtainable to ordinary loams, 
and in refractoriness from the highest testing fire clay to the ordinary 
glacial clays with a high content of lime and other impurities causing 
them to fuse readily. 

The sands and loams used in the foundry for molding purposes may 
be divided roughly into classes, according to the function which they 
perform, as follows: (1) sands used for making the cavities or holes in 
the metal castings; (2) sands or loams used to make the mold or container 
which gives the casting its exterior form. The former are called core 
sands, and the latter are called molding sands or loams. 

CORE SAND 

For core work a clean sand free from clay is preferred. To this clean 
sand is added some organic bonding material such as glue, resin, linseed 
oil, molasses, etc. The heat of the metal will destroy the bond, leaving 
the core sand incoherent so that it may be readily removed from the 
cavities of the casting. Greater permeability is obtained in this manner 
than is possible where loamy materials are used. 

The texture and quality of sand used for core work vary greatly and 
depend on the nature of the founding. If the cores are surrounded by 
large masses of metal and the temperature of the melt is high, a greater 
degree of permeability in the cores is required than if the metal mass is 
small and, the temperature low. Where a high degree of permeability is 
required, a sand of coarse texture in which little or no silty material is 
present is generally used. If the temperature of the melt is low or the 
mass of the casting is small, finer sands give satisfactory results for core 
work. 



NATURE OF SANDS AND LOAMS FOR FOUNDRY MOLDING 7 

Quartz sand entirely free from clay is preferred for core work, but 
it is seldom found in nature in commercial amounts, and sands with one 
and two per cent of clay are in common use. Glacial sands having a 
varied mineral content are also used. 

MOLDING SAND 

As generally defined a molding sand is a siliceous sand having a clay 
content just sufficient to bind the sand grains together, but not enough to 
fill the voids between the sand grains ; so that when moistened slightly 
the mass may be molded into any form desired, retaining this form when 
dry and being sufficiently open or permeable so that air may be blown 
through it. 

The chief requisites of molding sand, then, are permeability and ten- 
sile strength or ability to resist rupture. Molding sand must be permeable 
because when the hot liquid metal is poured into the sand mold consider- 
able volumes of steam and gas are generated in the sand. These must 
have the opportunity for ready escape, otherwise the mold will be blown 
to pieces by accumulated gas pressure before the metal has opportunity to 
set or cool. Molding sand must have sufficient bond so that the walls of 
the mold will stand up against the pressure of the liquid metal, which in 
large castings is considerable, and it must withstand the wash of the liquid 
metal as it is poured into the mold. 

Permeability and tensile strength requirements differ, consequently 
there is no exact standard to which all must conform. In certain classes 
of founding large volumes of gas are generated in the sand and a high 
degree of permeability is required, whereas in other classes of work but 
small amounts are generated and much less permeability is satisfactory. 
In the same manner the tensile strength or bonding power required of 
molding sand varies with the character of the founding. Accordingly a 
certain sand may be entirely satisfactory for one class of molding and 
unsuited for another. 

Other requisites of molding sand are certain textures and refractori- 
ness, but these also are measured by relative rather than absolute standards 
and depend on the class of founding done. For example, if the tempera- 
ture of the melt is high, say 3000 F., the molding sand can have but a 
small content of lime, alkalies, iron oxide, and other fluxing impurities ; 
otherwise the sand in contact with the metal will be fused or melted and 
will stick to the casting. For such work a refractory molding sand is 
essential. If, on the other hand, the temperature of the melt is less than 
1000 F., as in most brass work, there is little likelihood that the sand 
will be fused, and refractoriness as a property of such molding sand is 
for practical purposes negligible. 



8 THE FOUNDRY SANDS OF MINNESOTA 

The texture required of molding sand also varies with the class of 
founding. Some metals when melted are much more fluid than others 
and the fluidity of the same metal increases as the temperature is raised 
above the melting point. If the molding sand is very open and porous 
and the melt is of a high degree of fluidity, the liquid metal will pene- 
trate the sand or "search the sand," as the molders say. If the sand is 
exceptionally coarse in texture the face of the mold will be pitted and 
the surface of the casting made will be rough. For certain types of metal 
and for certain requirements of finished surface, molding sands of finer 
texture are required. 

Since a molding sand consists of a mixture of sand and clay it may 
be made artificially by mixing clean sand with the required amount of 
clay to give the desired bond. In fact, some foundries prefer molding- 
sand made in this manner. But in nature, sands and clays occur ready 
mixed in every conceivable proportion and most of the molding sands 
and molding loams are these natural mixtures. 

There are no simple tests and no infallible rules or formulae by which 
the inexperienced may judge whether or not a loam or sand may be suit- 
able for molding purposes but the skilled molder can usually judge the 
suitability of a given sand or loam for molding purposes by simply knead- 
ing it in his hands and by blowing air through it after molding it. The 
only satisfactory way for the layman to determine the probable value 
of such material for foundry use is to submit samples to experienced 
molders or foundrymen, or have the same subjected to laboratory tests. 

"DOCTORING" MOLDING SANDS 
It is well known that sands and loams as they occur in nature are 
rarely uniform in character. In any sand bank or pit the material usually 
varies in character from the surface downward, as well as laterally ; so 
that even if a sample of material is found having the right constitution 
for molding, there is no assurance that any considerable deposit will be 
found of exactly the same character. In actual foundry practice varia- 
tions in the sand or loam are anticipated and the material is brought to 
the desired consistency by adding the constituents that are lacking. If 
the loam is too heavy and lacks permeability, clean sand is added ; if it 
it too sandy and lacks bonding power, clay or heavier loam is mixed 
with it. Coal screenings or ground coal, called "sea coal," is sometimes 
added. Sawdust, chopped straw, and other similar organic materials are 
commonly used to give the desired texture and increase the permeability. 
Organic materials, such as glue, molasses, flour, stale beer, resin, linseed 
oil, glucose, etc., are frequently employed to increase the bond. This 
practice of modifying the consistency by adding certain constituents is 
called "doctoring" the sand. 



OUTLINE OF GEOLOGY OF MINNESOTA 9 

After the sand has been used in the mold one or more times, the bond 
is destroyed to a greater or less extent by the heat and the burned products 
accumulate in the sand as fine silty material, decreasing the permeability 
and changing the texture. Before being used again, the sand is doctored 
by adding such materials as will bring it back to the desired consistency. 
Thus the molding sand, which is continually deteriorating with repeated 
use, is restored to the proper consistency as to texture, bond, and per- 
meability by repeated doctoring which may consist in adding a single 
constituent or several constituents. In the ordinary foundry doctoring 
the sand is a part of the daily or hourly routine and taxes the skill and 
judgment of the molder to the limit. The grade of casting produced in 
a foundry, so far as soundness and exterior finish are concerned, depends 
largely upon doctoring to keep the sand to a uniform consistency. 

OUTLINE OF GEOLOGY OF MINNESOTA 

The materials used in the foundries include a rather wide range of 
products of rock weathering, i.e., different types of sands, loams, and 
clays that have been redistributed and redeposited by wind, glaciers, 
rivers, lakes, and the sea, and as now found are frequently far from 
their original sources. In view of this it was believed that the origin of 
this foundry material might be better understood and its variation in 
character better appreciated if a description of the material itself was 
prefaced by a brief account of the general geological relations of the 
state, the rock formations present, and the sequence of events leading 
up to the present disposition of the surface materials. In the following 
geological description the writer has borrowed freely from previous pub- 
lished reports on Minnesota geology. 

All of Minnsota except the extreme southeast corner is a broad gla- 
ciated plain differing only in minor details from the adjacent regions to 
the west, north, and east. The minor topographic features of this great 
plain are the result of glaciation and stream erosion that has taken place 
in postglacial time. They consist of : 

(a) Numerous belts of low knolls or gravelly hills, known as ter- 
minal moraines in which are innumerable small undrained basins, kettle 
holes, or ponds. These moraines represent the material that accumulated 
at the margin of the glacier in various stages of its advance and retreat 
and consist of a mixture of all types of material that was picked up from 
the local rock formations over which it passed. This morainic material, 
known as "glacial till," while exceedingly heterogeneous in mineral con- 
tent and physical character, at many places contains material that can 
be used in foundry work, as the detailed description shows. 



io THE FOUNDRY SANDS OF MINNESOTA 

(b) Broad, gently sloping plains or prairies of more or less well- 
assorted gravels and sand, known as outwash, usually fronting the ter- 
minal moraines, or otherwise definitely related to them, and representing 
the material carried out from the ice margin by glacial waters. These 
gravel plains are usually mantled by loam a few feet in depth, which is 
serviceable in foundry work, but the sands beneath rarely are. 

(c) Broad, comparatively flat tracts in which occur occasional low 
hills or knolls and many shallow ponds or lakes. These tracts are un- 
derlain by glacial till, known as ground moraine. The surface mantle 
of loam so persistent throughout the state is usually, though not always, 
present in these areas but is not ordinarily so well suited to foundry pur- 
poses as the loams otherwise disposed. 

(d) Long, gently sloping plains, usually narrow, that follow the pres- 
ent valleys or the lines of drainage in glacial time. They are underlain 
by gravels and sands more or less assorted and stratified. Some of these 
are similar in origin to the outwash plains and frequently are only con- 
tinuations of such plains, whereas others are postglacial in age and rep- 
resent the material redistributed by streams since glacial time. They are 
sometimes called valley trains. The usual mantle of surface loam is 
common in these areas and often affords good foundry loam. 

(e) Narrow plains of gravel and sand more or less well assorted 
that skirt the border of extinct lakes of glacial age. They seldom carry 
material of value in foundry work. 

The glacial material varies in thickness from o to 500 feet. If it 
were spread uniformly over the area where now found, it would be less 
than 100 feet thick. It is in reality only a veneer that mantles the to- 
pography of pre-glacial time and serves merely to mask the minor details 
of that ancient surface in some places and to conceal them entirely in 
others. In a major way Minnesota was a broad low plain in pre-glacial 
time, on which were well-established lines of drainage with broad shal- 
low valleys and the usual minor topographic features characteristic of 
stream erosion. The bed rock of the old formations appears at the sur- 
face at innumerable places in the glaciated area where the glacial drift 
was originally thin or absent and where subsequent stream erosion has 
uncovered it. 

At present Minnesota stands somewhat higher than in pre-glacial time. 
The southern portion of the state, considered broadly, is a low plateau 
with one elevated area in the southwest corner of the state, known as 
the Coteau des Prairie, rising to an altitude of 1,900 feet, or about 500 
feet above the general level, and a second domed or elevated area in the 
southeast corner, rising to 1,400 feet above sea level, which is about 400 
feet above the general level. 






OUTLINE OF GEOLOGY OF MINNESOTA n 

Northern Minnesota possesses a more varied topography. The north- 
western part is as flat as a floor, except for a series of low ridges, being 
the site of the ancient Lake Agassiz. The extreme northeast part of 
the state is more rugged with several low "ranges" rising to altitudes of 
i, 800 to 2,000 feet. 

The southeast corner of the state, representing about one sixteenth 

of its total area, is altogether different from the rest of the state in 

topographic appearance. In a portion of this area, in Houston and 

Winona counties, glacial drift is absent and probably was never present. 

It is a part of the so-called "driftless area" of the upper Mississippi Val- 

I ley. Here natural lakes and ponds so abundant in the rest of the state 

; are entirely absent and the surface everywhere slopes toward the streams 

I that drain it. 

Bordering the Mississippi River in this area the tributary streams 
1 coming from the west have cut valleys 200 to 500 feet deep, which in 
1 places are gorgelike and in other places are flat-bottomed, walled in by 
1 cliffs of massive limestones and sandstones of the old Paleozoic forma- 
i tions. Between the valleys the upland areas are commonly rather flat 
1 or very gently sloping, plateau-like in aspect, and are remnants of an 
1 old plain of stream degradation developed in pre-glacial time when the 
I land stood much lower than now. 

The uplands in this area are more or less completely covered by a 

siltlike loam called loess, to be described later, o to 40 feet in depth, which 

1 mantles the high points and reaches down the valley slopes even to the 

streams in places. The old rocks native to the region are largely buried 

1 by this loess but in the steeper slopes they protrude and exposures are 

seen in abundance. 

The rocks of Minnesota 2 range in age from the most ancient known 

as the Archean to the youngest which are now in the process of deposi- 

I tion along the present streams and in the lakes and are known as recent. 

The accompanying tabulation shows the principal formations present in 

1 the state arranged in their proper sequence with the oldest at the bottom. 

1 It will be observed that there are two breaks in Minnesota's stratigraphic 

2 Winchell, N. H., Geology of Minnesota: Final Report Geol. and Nat. Hist. Survey of 

I Minnesota, Vols, i to 5, 1882-1900. 

Leverett, F., Surface formations and agricultural conditions of northwestern Minnesota: 

II Minnesota Geol. Survey Bull. 12, pp. 1-75, 1915. 

Leverett, F., and Sardeson, F. W., Surface formations and agricultural conditions of north- 
t eastern Minnesota: Minnesota Geol. Survey Bull. 13, pp. 1-72, 1917. 

Leverett, F., and Sardeson, F. W., Surface formations and agricultural conditions of 
! southern Minnesota: Minnesota Geol. Survey Bull. 14, pp. 1-178, 1919. 

Grout, F. F., and Soper, E. K., Preliminary report on the clays and shales of Minnesota: 
1 Minnesota Geological Survey Bull. 11, pp. 1-1 72, 19 14. 

Grout, F. F., and Soper, E. K., Clays and shales of Minnesota: U.S. Geol. Survey Bull. 678, 
1 pp. 1-259, 1919. 

Bowles, O., Structural and ornamental stones of Minnesota: U.S. Geol. Survey Bull. 663, 
pp. 1-225, 1918. 



TABLE I. GEOLOGIC COLUMN OF MINNESOTA 



Era 


System and Series 


Formation 


Approximate 
Thickness 


Character of Strata 


Value for 
Foundry Use 


CENO- 
ZOIC 


Recent 


Recent 


o to 300 


Sands, silts, clays, 
muds 


Little or none 
Valuable 


Pleistocene 


Glacial 


to 600 


Loess, gravel, sand, 
loams, clays 


Loams valuable, 
sands less so 


do 
[AS 


Cretaceous 


Benton shale 
Dakota formation 


to 550 


Clays and sriales 
Sands, clays 


Clays valuable, 
sands less so 






Devonian 




100 


Sandstones and 
limestones 


No value 




Ordovician 


Maquoketa 


100 


Shale and 
limestones 


No value 




Galena 1 
Decorah [ 
Platteville J 


230 


Limestone 1 
Shale \ 
Limestone J 


No value 


u 

o 

N 


St. Peter 


80-200 


Sandstone 


Valuable quartz 
sand 


o 


Shakopee 


100 


Dolomite 


Small value 


h-3 

0< 


Oneota 


75-200 


Dolomite 


Small value 


Cambrian 


Jordan 


75-200 


Sandstone 


Most valuable 
quartz sand 
in state 




St'. Lawrence 


100-200 


Dolomite and shale 


No value 




Franconia 


50-100 


Sandstone 


Little or no value 




Dresbach 


300-450 


Sandstone, shale, 
and limestone 


Little or no value 




Algonkian 


Red Clastic series 
in southern Minne- 
sota, sandstone in 
northern Minne- 
sota 


2,250 
maximum 


Sandstones and 
shales 


Kettle River sand- 
stone valuable for 
quartz sand 
Other sand- 
stones also 




Keweenawan 


Conglomerate and 
sandstone 


500 


Conglomerate and 
sandstone 


Little or no value 




Eruptives 


Unknown 


Igneous rocks 


No value 


o 

t— 1 

o 

N 

o 


< 

o 


< 

1— 1 

o 


Upper 
Huronian 
(Animikie 
Group) 


Intrusives 


Unknown 


Acid and basic 
igneous rocks 


No value 


Virginia and 
other slates 


3,000 


Slates 


No value 


w 

H 

o 
Hi 


Biwabik and 
Gunflint 


800 


Taconite, iron ore, 
chert 


No value 


Pokegama 


200 


Quartzite, Sioux 
quartzit'e 


Valuable for 
quartz sand 




Lower 

Middle 

Huronian 


Giant Range 


Unknown 


Granite, dolomite 
porphyries 


No value 




Graywacke 


5,ooo 


Slate, graywacke, 
conglomerate 


- i, 

No value 




w 
W 
u 

< 


Laurent'ian 


Residual clay 




Residual clay 


- 
Possibly valuable 




Igneous rocks 


Unknown 


Granites, schists, 
porphyries 


No value 




Keewatin 


Soudan 


Unknown 


Banded cherts 

and jaspers, iron ore 


No value 




Ely and other 
formations 


Unknown 


Greenstone, schists, 
porphyries 


No value 



OUTLINE OF GEOLOGY OF MINNESOTA 13 

record. One occurs between the Ordovician and Devonian, the Silurian 
rocks being absent ; and a second occurs following the Devonian, the 
Carboniferous, Permian, Jurassic, and Triassic rocks being absent. The 
Cretaceous is found lying directly on the Devonian and older formations. 

Archean period. — The rocks of the Archean system in Minnesota are 
divided into two groups, i.e., the Keewatin group at the base and the 
Laurentian group above. The Keewatin epoch was one of vulcanism and 
the rocks of the Keewatin group consist largely of lava flows that are 
for the most part basic. The land evidently stood low for water-laid 
sediments were deposited alternating with the lava flows. 

The entire series of beds was subsequently complexly folded and 
profoundly metamorphosed converting the basalts into green schists now 
seen in outcrop in much of the northern part of the state. 

The Laurentian epoch resembled the Keewatin in that igneous activity 
was dominant, but the igneous material was intruded instead of being 
extruded as lava flows. 

The rocks of the Archean period in general are of little value for 
foundry purposes unless it shall be found that some of the more re- 
fractory schists may be serviceable for furnace lining. 

Algonkian period. — The rocks of the Algonkian period are divided 
into three groups which from the base upward are known as the Lower- 
Middle Huronian, the Upper Huronian, and the Keweenawan. 

As compared with the Archean the Algonkian period was one of sedi- 
mentation, but pronounced igneous activity occurred at intervals during 
which times igneous material both basic and acid and of several types was 
intruded in the stratigraphic series. 

At the close of the Middle Huronian epoch the region was lifted 
above sea level and more or less erosion occurred, resulting in an un- 
conformity between the Middle and the Upper Huronian groups. 

The water-laid sediments of the Algonkian period consisted of gravels, 
sands, and muds which as the result of complex alterations by both chem- 
ical and dynamic metamorphism were changed to conglomerates, sand- 
stones, quartzites, shales, and slates. Some of these formations, as will 
appear from the more detailed description, yield foundry sand, particu- 
larly the quartzite at New Ulm and the Sioux quartzite (pp. 18 and 19). 

The Red Clastic series. — Above the Keweenawan beds there appears a 
group of beds known as the Red Clastic series, about 2,250 feet in maxi- 
mum thickness, consisting of shales and sandstones, which are provi- 
sionally referred to the Algonkian. Locally this group carries sandstones 
of exceptionally high quartz content such as the Kettle River and Hinckley 
sandstone and certain sandstones in vicinity of Duluth which furnish 
good sand for refractory purposes both for foundry and for steel fur- 
nace use. 



i 4 THE FOUNDRY SANDS OF MINNESOTA 

Cambrian period. — At the close of the Algonkian period all of Min- 
nesota stood above sea level and there was a prolonged period of subaerial 
erosion and decay of the old crystalline rocks. The region remained in 
this relation through the greater part of the Cambrian period and not 
until late Cambrian time did it subside, permitting the sea to encroach 
from the south and west, resulting in the deposition of sand, sandy muds, 
and sandy limestones over the greater part of southern Minnesota. The 
northern part of the state probably remained above sea level throughout 
Cambrian time and continued to suffer erosion instead of deposition. 

The Cambrian group of beds in Minnesota consists of four formations 
which from the bottom upward are known as the Dresbach, Franconia, 
St. Lawrence, and Jordan. The Cambrian sea was shallow throughout 
the period and the long period of subaerial decay preceding the Cambrian 
submergence had prepared a vast amount of soft and loose material for 
redistribution by the shallow Cambrian waters. This is evidenced by 
the incomplete sorting of the material in the Dresbach and Franconia 
formations, and the rapid alternation of thin beds of fine sand and sandy 
mud in these formations. 

In the last epoch of the Cambrian period, i.e., the Jordan, conditions 
were somewhat different, either as to supply of material or as to depth 
of the sea, with the result that a quartz sand of exceptional purity was 
deposited over a wide area in the southern part of the state. 

There was little if any orographic movement during the Cambrian 
period and there has been but little since, so that the Cambrian beds 
have hardly been disturbed and very slightly altered in most places. The 
Cambrian beds lie in nearly horizontal position much as they were de- 
posited and in many localities are so slightly cemented that they may be 
mined with steam shovel or by hand without the aid of blasting. 

The Cambrian formations carry many beds of sand that locally are 
sufficiently well assorted and of sufficiently clean quartz to be valuable 
for foundry purposes; but the uppermost member, i.e., the Jordan, so 
far exceeds the other formations in the quality and abundance of the 
sand that the other Cambrian formations are hardly worth special 
consideration. 

Ordovician period. — There was apparently no elevation of the region 
at the end of Cambrian time — -on the contrary, probable further subsid- 
ence — and deposition continued without interruption into the Ordovician 
period. The first deposits were calcareous muds now recognized as the 
Oneota and Shakopee dolomites through which appear numerous thin 
beds and lenses of quartz sand often remarkably clean and not unlike 
the Jordan. 



OUTLINE OF GEOLOGY OF MINNESOTA 15 

After the deposition of these calcareous muds of the Shakopee, there 
was a slight elevation of the region, shallow water conditions prevailed 
over a remarkably wide area and a quartz sand, known as the St. Peter, 
was spread over all of southern Minnesota and adjacent parts of Wis- 
consin and Iowa, reaching far southward into Illinois, Missouri, and 
beyond. For uniformity of material and areal extent, this is one of the 
most remarkable formations in the United States. About the continental 
border to the northward the slight elevation at the beginning of St. Peter 
time probably brought the Jordan formation above sea level. The Jordan 
here was eroded and redistributed, thus contributing in part to the St. 
Peter deposition ; but the original crystalline rocks to the northward, 
deeply decomposed, were probably the chief source of this great supply 
of quartz sand. 

Following the St. Peter epoch there was a slight subsidence of the 
region and the recurrence of conditions favorable to the alternate deposi- 
tion of limestones, muds, and clays which are now recognized as the 
Platteville limestone, the Decorah shale, the Galena limestone, and the 
Maquoketa shale. The shale formations in this series afford much clay 
which is being used extensively in the manufacture of brick, terra cotta, 
tile, etc., but because of the large content of fluxing impurities these 
clays are of little interest in foundry work. The Ordovician series of 
beds is of chief interest to the foundry industry for the supply of quartz 
sand afforded by the St. Peter formation described in more detail in 
the following pages. 

Devonian period. — With the close of the Ordovician period the region 
was again elevated and remained above sea level throughout the Silurian 
period, after which subsidence again occurred sufficient to allow the sea 
to encroach upon the southern part of the state, causing the deposit of 
the Devonian limestone 50 to 100 feetf thick, together with some shale. 

Cretaceous period. — Following the Devonian deposition the region 
was again elevated and remained above sea level for a very long time, 
i.e., throughout the Carboniferous, Permian, Jurassic, and Triassic pe- 
riods. 

The elevation was apparently slight and not much erosion was ac- 
complished. The relations, however, were such as to permit the 
decomposition of the old crystalline rocks to considerable depths ; the 
feldspars were changed to kaolin or clay in which the quartz and other 
resistant mineral remained embedded, the sedimentary rocks such as 
sandstones and quartzites disintegrated into sand, and the limestones 
were leached, leaving a clay residuary. Thus there accumulated a great 
amount of loose or soft material ready for redistribution when later the 
region again subsided and the Cretaceous sea encroached upon the land. 



i6 THE FOUNDRY SANDS OF MINNESOTA 

This sea encroached upon Minnesota from the south and west re-work- 
ing the accumulated products of rock, weathering and redistributing them, 
depositing sands and clays with some limestone upon the upturned edges 
of the older eroded formations. It is thought the Cretaceous deposition 
reached as far north and east as the Mesaba Range. 

After the Cretaceous period the region was again elevated and stream 
erosion carried away much of the material that had been deposited but 
considerable areas in the southwestern part of the state remained, and 
detached areas or erosion remnants of greater or less extent are found 
in Other parts of the state. These afford valuable supplies of semi- 
refractory clay and some sand of importance in the foundry industry, 
as appears in the more detailed description of these deposits which follows. 

Pleistocene. 3 — Just before the Pleistocene epoch began the larger 
features of the topography of Minnesota were much the same as now. 
The advance of the ice, however, scoured out some valleys, filled in others, 
and deposited drift over large areas, entirely changing the details of 
topography and drainage. The retreat of the ice, by melting, formed 
marginal lakes and swollen, heavily silted streams, with further deposi- 
tion and modification of topography. 

The Pleistocene deposits show a peculiarly complex history, recording 
not only recurring stages of glaciation separated by long stages of de- 
gradation but also a complexity of ice movement within a single glacial 
stage. In the latest stage (known as the Wisconsin stage) there was one 
movement into Minnesota from the northwest, another from the north, 
and a third from the northeast ; and possibly there was similar complexity 
in earlier stages. These movements were not synchronous in their ad- 
vance, culmination, and waning, but each had its time of waxing and 
waning. 

The oldest glacial deposit, known as the pre-Kansan or Nebraskan 
drift, is almost completely buried beneath later deposits. The attenuated 
edge of this drift may be exposed outside the Kansan drift in the south- 
eastern counties. 

Clayey and silty deposits found in a few places under the Kansan 
drift may represent accumulation in the Aftonian interglacial stage. 
They are, however, thin and of small extent. 

The Kansan drift is extensively exposed in the southeastern part of 
Minnesota and in Pipestone and Rock counties in the southwestern part. 
It is generally of clayey texture, but as a rule carries limestone pebbles. 
It contains local pockets or lenses of marly pebbleless clay. 

Weathering has somewhat modified the character of the upper part 
of the Kansan drift and wind has coated much of the surface in the 

3 Leverett, F., Clays and shales of Minnesota: U.S. Geol. Survey, Bull. 678, p. jj, 1919. 






FORMATIONS YIELDING FOUNDRY MATERIALS 17 

southwest corner of the state and some of it in the southeast part with 
several feet of loess, a pebbleless deposit of fine dust. Recent wash has 
moved some loess and some drift down sloping hillsides and so mixed 
them that the precise mode of origin of some particular clay banks may 
not be easily determined. 

The Wisconsin gray drift, which was deposited by an ice sheet mov- 
ing from Manitoba across western and southern Minnesota, is largely of 
clayey texture, but, like the Kansan drift, contains many pebbles, a large 
part of which are limestone. In places some lenses and pockets of pebble- 
less clayey material are included in the stony clay. 

The Wisconsin drift that was deposited by glaciers invading eastern 
and central Minnesota from the north and northeast is in large part 
stony and sandy and contains very few bodies of silt or clay. 

Locally streams of water from the melting ice spread fine sandy clays 
over what are called outwash plains. 

The lake silts laid down in bodies of water in the western part of 
the Lake Superior basin and in the Red River basin are in places very 
thick, as at Wrenshall, but generally there is only a thin deposit of lacus- 
trine sediment in the deep pools of these old lake beds and in certain 
localities there is only sand. 

On the borders of the Lake Superior basin red boulder clay was de- 
posited in large quantities ; it is now dissected by steep gorges. 

In addition to the great glacial lakes of the Lake Superior and Red 
River basins, there were numerous small lakes or ponded areas along 
the border of the ice, in which silt was deposited. Such an area lies north 
of Princeton at the large plants of Brickton and similar 1 areas lie farther 
east along the northern border of a district which was occupied by a 
lobe of ice that extended northeastward from the Mississippi to the St. 
Croix Valley, above the Twin Cities. These clays were, in places, over- 
ridden by the later advances of the ice. 

Recent. — Since the final disappearance of the Pleistocene glaciers there 
have been only slight modifications of the surface and very little accumu- 
lation of material or removal by erosion. A few lakes and swamps have 
been drained, and a few have been filled. The large river channels have 
been covered with silt. 

GEOLOGICAL FORMATIONS YIELDING FOUNDRY SANDS, 

LOAMS, AND CLAYS 

The Sioux quartzite formation. — The oldest geological formation of 
importance to the sand industry is the Sioux quartzite which belongs to 
the Upper Huronian group, in the Huronian system. 



i8 THE FOUNDRY SANDS OF MINNESOTA 

This quartzite is exposed at numerous places in the state but because 
of its hardness it is expensive to quarry and is of little service in foundry 
work except when crushed. This quartzite is extensively quarried at 
Courtland on the Minnesota River ; in Nicollet County, sometimes called ] 
the New Ulm quartzite; also at Pipestone, Jasper, and Luverne in the 
southwest corner of the state, and at Sioux Falls, South Dakota, where 
it is known as the Sioux quartzite. 

By reason of its extreme hardness, density, and toughness, this quartz- 
ite withstands attrition and abrasion to a remarkable degree and is much 
in demand as "pebbles" for ball mills. The quarries at Jasper supply the 
trade. The quartzite is cut into rectangular blocks about 3x3x4 inches, 
which are rounded to pebble shape by tumbling in revolving cylinders, 
and are shipped for use in ball mills. 

The defective blocks and rubble of the quarry are crushed for con- 
crete aggregate and for road material, and in the process of crushing 
more or less fine material results, which is screened out and sold as sand, 
the general texture of which is shown in the mechanical analyses Nos. 
200, 205, and 207, Table XIV. This sand, screened to the proper size, 
should be very effective as a blast sand for foundry use and should be 
valuable also for refractory purposes as the following determination of 
the mineral composition shows 4 : 

Quartz, (Si0 2 ), 96.3 % ; Orthoclase (KA1 Si 3 O s ), 1.4 % 
Zircon, (Zr Si0 4 ), .1 % ; Hematite (Fe 2 O s ), 2.2% 

It probably would not be feasible to crush the Sioux or New Ulm 
quartzite for fine screenings or sand, but where such screenings may be 
obtained as a by-product at a nominal cost they are worthy of considera- 
tion for foundry use both as refractory sand for furnace linings and as 
blast sand. It should be noted that it is quartzite of this type that is 
elsewhere crushed and used in the manufacture of "ganister" brick for 
furnace lining, a small amount of lime or other fluxing material being 
used to insure incipient fusion to bind the material when burned. 

The Kettle River formation. — In the vicinity of Hinckley, Pine 
County, midway between St. Paul and Duluth there is a quartzitic sand- 
stone known as the Kettle River or the Hinckley sandstone which under- 
lies a considerable area. It is concealed by a thin mantle only of glacial 
drift and the streams of the region have uncovered it at numerous points. 
Kettle River has cut a narrow canyon through it 50 to 100 feet deep 
for a distance of several miles. 

At the town of Sandstone, on Kettle River, extensive quarries have 
been opened in the formation, for the rock is a high-class building stone, 
and until a few years ago when concrete so largely replaced stone for 
structural purposes, was much in demand and has been used in numerous 

4 Bowles, O., op. cit., p. 202. 



FORMATIONS YIELDING FOUNDRY MATERIALS 19 

buildings in Minnesota, Iowa, Illinois, and South Dakota. Laboratory 
tests show that this Kettle River sandstone used as aggregate makes a 
superior concrete and there is at present a growing demand for it for 
this purpose so that the output of the quarries at present is largely crushed 
rock. 

In density this rock is about midway between the Sioux quartzite and 
ordinary sandstone. The cementing material is principally silica and the 
quartz grains of the original rock have to a large degree been enlarged 
by secondary quartz showing crystal faces. The rock, however, is still 
granular and upon crushing breaks along the faces of the grains instead 
of through the grains themselves. The screenings that result from crush- 
ing are more nearly true sand than are those from the Sioux quartzite, 
and a greater volume is obtained. These screenings are sold as sand 
and have been used for refractory purposes in the foundries of Minne- 
apolis and St. Paul and in the steel furnaces of Duluth with satisfactory 
results. The subangular shape of the grains and their roughened sur- 
faces offer good attachment for the bonding material which is an ad- 
vantage for certain purposes. The comparative mesh of the sand, as 
shown by 185B, Table X, is 58, over 35 per cent being coarser than 60- 
mesh sand. 

Chemical analyses of the Kettle River sandstone show the, following- 
composition : 

No. 1 No. 2 

Silica, Si0 2 97.10 98.69 

Alumina, AL0 3 2.20 1.06 

Lime, CaO 60 .42 

Magnesia, MgO 10 .01 

Potash, K 2 trace 

Soda, Na.O .17 



100.00 100.35 

No. 1. Analysis by United States arsenal, Watertown, Mass. 

No. 2. Winchell, N. H. Geology of Minnesota, Vol. 1, p. 202. 1884. 

The Jordan formation. — -The Jordan formation which is the uppermost 
member of the Cambrian group underlies all of southeastern Minnesota, 
but being overlain by several younger formations, it is deeply buried in 
the greater part of the area and is known to be present only from the 
records of deep wells. Natural exposures of the formation are seen, how- 
ever, along the northwest border of the area. There are exposures along 
the valley of the Minnesota River and along the Mississippi below Red 
Wing where such tributaries of the Mississippi as Zumbro and Rock rivers 
have cut into the Cambrian beds. 



20 THE FOUNDRY SANDS OF MINNESOTA 

The only pits where the Jordan was being mined for sand on a com- 
mercial scale were found in the valley of Minnesota River in the vicinity 
of Ottawa and Kasota, Le Sueur County. The pits in these localities were 
one-half to one mile from the railroad stations and all are between the 
railroad and the river. Numerous other exposures where good sand might 
be obtained from the Jordan were seen between Ottawa and Merriam 
Junction, thirty miles to the northeast, but no pits had been opened. There 
was little inducement to open new pits because the pits at Ottawa and 
Kasota are almost ideally located. They are close to the railroad, have 
excellent drainage, and the banks of sand are 10 to 20 feet deep. The sand 
is so slightly coherent that a steam shovel might be operated without 
blasting. Sand could be produced there in practically unlimited quantities 
at a nominal cost per ton. 

The southeastern part of the state in Wabasha, Winona, and Houston 
counties was not studied in sufficient detail to determine to what extent the 
Jordan is being used as a source of sand in this part of the state, or its 
availability. 

The samples of Jordan sand, collected in Fillmore County at Preston 
and in Wabasha County at Mazeppa, were so similar in texture and mineral 
content to those collected from the Jordan sandstone in Le Sueur County 
that it was difficult to distinguish between them. These samples, together 
with field observations generally, serve to show that the Jordan is remark- 
ably uniform in character over wide areas. Cursory examination showed 
it difficult to find anything but quartz in most of the samples. There can 
be little doubt that the Jordan furnishes sand as nearly pure quartz as 
any sand in the state. The formation which is 75 to 200 feet thick is not 
entirely uniform in character or in texture, but in all localities where good 
exposures of the formation were seen, beds 10 to 20 feet in thickness were 
found in which the sand was essentially pure quartz, with a texture such 
as shown in samples 8, 19, 170 etc., Table X. 

The sand of the Jordan formation, in its coarser phases at least, which 
are preferred for foundry work, consists of quartz grains which in the 
larger sizes, i.e., the 20 and 40 mesh, are exceptionally well rounded ; in 
fact occasional grains are found that are almost perfect spheres and even 
in the smaller sizes, i.e., 60 and 80 mesh, the grains are less angular than 
in average quartz sand. 

Standard sand, known to the engineering testing laboratories through- 
out the country as Ottawa sand, comes from Ottawa, Illinois, and consists 
of essentially pure quartz sand sized to 20 mesh. Sand of this size screened 
from the Jordan sand from Ottawa, Minnesota, or from Merriam Junction 
can not be distinguished from the standard sand of Ottawa, Illinois. 



FORMATIONS YIELDING FOUNDRY MATERIALS 21 

Other sand formations occur in the Cambrian group in Minnesota in 
the Franconia and Dresbach formations, but in uniformity of texture, 
mineral content, and in availability they do not compare favorably with 
the Jordan sand. 

The St. Peter formation. — The St. Peter formation appears in the 
geologic section about 200 feet above the Jordan formation and belongs 
to the next higher group or system, the Ordovician. 

The St. Peter like the Jordan consists essentially of quartz sand. It is 
even more widespread than the Jordan, extending eastward and southward 
far beyond the boundries of Minnesota. It is rarely cemented sufficiently 
to be of value as building stone and is on the whole finer than the Jordan. 
Certain horizons in the St. Peter formation locally are as coarse as the 
Jordan, but rarely are such horizons thick enough to warrant separating 
them from the finer beds above and below. 

In fresh exposures the St. Peter sand is usually very white and it 
has the appearance of exceptional purity but a bank sample of the material 
usually contains 1 to 2 per cent clay which is more than the Jordan carries. 
It also contains more silt than the Jordan. 

For ordinary core work in foundries the St. Peter is very satisfactory, 
and since it constitutes the bed rock underlying the greater part of Minne- 
apolis and St. Paul an ample supply of this sand is available in these cities. 
The St. Peter sand upon weathering takes on a yellow color due to iron 
staining and in this condition is sometimes preferred to the fresh white 
sand in foundry work. The preference for this type of sand is probably 
to be attributed to the slight roughing of the surface of the quartz grains 
by the deposit of iron oxide which affords a better attachment for the 
bonding material used. 

In the region from St. Paul southward to Northfield, a distance of 
forty miles, the St. Peter formation is generally the bed rock. The glacial 
drift being thin in much of this area, numerous exposures of the St. Peter 
appear, usually as a soft yellow sand, or as friable sandstone. From 
Northfield eastward to the Mississippi along the valley of Cannon River 
the St. Peter formation is conspicuously exposed in the flanks of the flat 
topped, mesa-like hills and in the valley sides as white or yellow sand. 

From Red Wing to the Iowa line the St. Peter is seen in the valleys 
tributary to the Mississippi at gradually lower and lower levels as one goes 
southward. The exposures are too numerous to warrant special descrip- 
tion. The most characteristic feature of the sand is its uniformity in 
texture. 

The Oneota and Shakopee dolomites, which occupy the interval be- 
tween the Jordan and the St. Peter formations, also carry thin beds or 
lenses of sand or sandstone similar to the Jordan or to the St. Peter, but 



22 THE FOUNDRY SANDS OF MINNESOTA 

such beds are usually pocket-like and rarely of commercial importance, 
except very locally. 

A chemical analysis 5 of the St. Peter sandstone from Fort Snelling 
is as follows : 

Silica, SiO a 97-67% 

Alumina, Al 2 O s i-3i% 

Lime, CaO 0.41% 

Potash, K 2 0.02% 

Iron, Fe 2 3 0.55% 

Soda, Na 2 0.15% 

A bank sample of the St. Peter sand, such as is used in foundry work, would probably show 
a considerably lower silica content, but samples from selected horizons of the clearer sand show 
a silica content of 98 per cent. 

The Decorah formation. — The Platteville limestone, Decorah shale, 
Galena limestone, and Maquoketa shale formations overlie the St. Peter 
sandstone and occupy the upper part of the Ordovician group of beds. 
The Decorah shale is extensively used in different parts of the state for 
making* brick, tile, and terra cotta. It is of minor importance as a source 
of clay for foundry use because of the rather large percentage of fluxes 
usually present. 

The Cretaceous formation. — As explained in the foregoing geological 
history a long interval of subaerial erosion and weathering immediately 
preceded the Cretaceous period, during which time the old crystalline 
rocks were deeply decomposed, the sedimentary rocks disintegrated, and 
the limestones leached of their readily soluble constituents. The feld- 
spars of the crystalline rocks were kaolinized and in turn leached of much 
of their more soluble lime and alkalies leaving semi-refractory clays as a 
residuary. 

The Cretaceous sea encroaching upon the land found great quantities 
of soft and loose products of rock-weathering ready for redistribution, 
and accordingly the deposits of the Cretaceous period consist of a great 
variety of materials derived from the acid and basic crystalline rocks 
and from the sandstones, limestones, and shales of the sedimentary 
formations. 

In re-working this material the Cretaceous sea in some areas accom- 
plished a remarkably complete separation of the coarse and fine materials, 
depositing beds of dense fat clays in certain localities, or at certain hori- 
zons, and clean, well-assorted sands in others. In other areas the sepa- 
ration was less complete or very poor and the products of rock-weathering 
were redeposited as poorly assorted clays and sands, or as alternating- 
thin beds or laminations of these materials. 

5 Winchell, N. H., Geology of Minnesota Final Report, Vol. 1, p. 202, 1884. 



FORMATIONS YIELDING FOUNDRY MATERIALS 



n 



The Cretaceous formation in general has suffered little alteration or 
induration since its deposition, so that the beds are still found in much 
the same condition as they were deposited. The sands are commonly 
loose, incoherent material, and friable sandstone. The argillaceous ma- 
terial appears as clays or soft shales. 

The old rock floor on which the Cretaceous was deposited was marked 
in certain localities at least, by numerous small potholes and caves, in 
which the clays accumulated in the manner illustrated in Figure i. 




\ >— » - ''— -, 



FIGURE I. SECTION OF CRETACEOUS CLAY AT SOUTH BEND, (a) SHAKOPEE 

dolomite; (b) cretaceous clay; (c) drift 

AFTER N. H. WINCHELL 



One of the important deposits of foundry clay worked in the vicinity 
of Ottawa, Minnesota, occurs in these relations in potholes in the basal 
portion of the Oneota limestone, and in the top of the Jordan formation, 
samples of which appear in Table XIII. 

One of the largest deposits of Cretaceous clay in the state, at present 
worked as a source of clay for the Red Wing potteries, is found in Good- 
hue County in the vicinity of Bellechester and Clay Bank. This clay 
has been used to a limited extent only in foundry work, but merits more 
extended use, since it is semi-refractory and certain beds in the series 
are highly plastic. The clay here occurs as thin beds for the most part 
alternating with beds of sand. 

The sands of the Cretaceous frequently consist of nearly pure quartz, 
but such beds are usually thin, and there is also a considerable variation 
in texture in the different horizons. In view of the fact that much bet- 
ter sand for foundry purposes is obtainable from the St. Peter and the 
Jordan formations, the Cretaceous formation does not warrant special 
consideration as a source of quartz sand. 

The Cretaceous formation is the source of the most refractory clay 
in the state. Chemical analyses of these clays show the following com- 
position. The localities from which the foundry samples were collected 
are the same as for those given in the following tables. 



24 



THE FOUNDRY SANDS OF MINNESOTA 



TABLE II. ANALYSES OF BUFF TO WHITE BURNING CRETACEOUS SHALES 



Silica 

Alumina 

Ferric oxide 
Ferrous oxide . . 

Lime 

Magnesia 

Sodium oxide . . . 
Potassium oxide . 
Phosphorus oxide 
Sulphur trioxide 

Titanium 

Moisture 

Ignition 

Total 



69.92 

17.39 

1.68 

.60 
1. 11 

.07 
2.25 

.63 
1. 10 

5-45 



69.84 
23-07 



.11 

.14 

[Trace 



6-35 



99.99 



68.298 
18.266 

2.867 

.719 
.802 
.81 
.60 



1.29 
6.155 



69.050 
18.830 

2.607 

.296 

.622 
1.066 
1. 461 



4.912 



99.807 99-742 101.39 





59 


72 


30 


00 




82 






51 




10 


34 



70.10 

16.99 

Trace 



10.69 
• 23 



99.99 



87.70 
7.24 

Trace 



Trace 
Trace 



99-34 



68.70 
18.04 

i-53 

1.24 
.56 
.24 

5.28 
.09 



1.40 



97 08 



1. Red Wing, Goodhue County. Clay sampled at stoneware plant. F. F. Grout, analyst. 

2. Red Wing, Goodhue County. Analysis reported by J. H. Rich to Heinrich Ries. 

3. Red Wing clay. Sample from Minnesota Stoneware Co., Red Wing, April 22, 1902. 
C. P. Berkey, analyst. 

4. Red Wing stoneware clay, air dried. C. P. Berkey, analyst. 

5. Ottawa. Ottawa Brick Co. Ries, Heinrich, Clays; their occurrence, properties, and 
uses, 1906. 6, 7, and 8. Minnesota Geol. and Nat. Hist. Survey Final Rept., Vol. 1, p. 438, 1884. 

6. Near Mankato. Clay filling hollows in Shakopee dolomite. 

7. Near Mankato (sec. 20). White clayey bed of considerable extent. 

8. Near Mankato. Clay or shale between Shakopee dolomite and Jordan sandstone in 
L'Huillier Mound. 

TABLE III. ANALYSES OF RED-BURNING CRETACEOUS SHALES 



Silica 

Alumina 

Ferric oxide . . . 
Ferrous oxide . . 

Lime 

Magnesia 

Soda 

Potash 

Barium oxide . . . 
Manganese oxide 
Phosphorus oxide 

Sulphur 

Titanium oxide . 

Moisture 

Loss on ignition 

Total 



63-65 
17.27 

4-75 

1. 21 

.06 

.91 

2-47 



.62 
2.03 
7.36 



100.33 



61.32 

12.27 
3.62 
4.18 

• 99 

1.76 
.42 

3-59 
.05 
.27 
.27 
.19 
.66 



10.73 



100.32 



58.14 
19.40 

5.52 



• 79 

1.52 

•54 

2.09 



.68 
2.10 
8.81 



99-59 



73-34 
14-75 

- 5-45 
.28 

.05 
Trace 
Trace 



4.71 



98.58 



1. Gray shale, west of Springfield. F. F. Grout, analyst. 

2. New Ulm, Minn. Brick clay, U.S. Geol. Survey Bull. 60, p. 151, 1890. T. M. Chatard, 
analyst. A brick made from this clay is reddish brown, strongly sintered, somewhat fractured. 
Sample taken by John Lind on south bank of Cottonwood River on section line at river crossing, 
east of wagon-road crossing, south of New Ulm. 

3. Clay sampled by A. Parker, of Brown Valley, Minn., just beyond the state line in South 
Dakota. F. F. Grout, analyst. 

4. Red ochery clay, near Mankato. Minnesota Geol. and Nat. Hist. Survey Final Rept., 
Vol. 1, p. 438, 1884. 



FORMATIONS YIELDING FOUNDRY MATERIALS 25 

As explained elsewhere, however, in the description of the foundry 
loams, the Cretaceous beds frequently consist of an intimate mixture of 
nearly pure quartz sand and semi-refractory clay, the deposits at Belle- 
chester being a case in point. For material of this type the Cretaceous 
merits consideration. 

Glacial deposits. — From what has been said above it will be readily 
understood that the glacial deposits are extremely variable in character, 
physically, mineralogically, and chemically, therefore it is difficult to de- 
scribe them briefly or to explain in what ways they are unsuited to cer- 
tain purposes and well adapted to other purposes. In each of the suc- 
ceeding periods of glaciation the ice picked up more or less of the older 
glacial material, mixed it with material newly derived, and redeposited 
it in great confusion. 

With reference to the sands and foundry materials the glacial de- 
posits of Minnesota may be grouped as follows : Moraines of Wisconsin 
drift from the north and northeast (red drift) ; moraines of Wisconsin 
from the northwest ; outwash ; and glacial lake deposits. These groups 
are based on differences in character and not strictly on origin, for ex- 
ample, the gray drift moraines include material deposited in two or three 
separate invasions of ice. 

Moraine deposits. — W T here the ice held its position for considerable 
time, the deposits laid down at the edge of the ice sheet are called moraines. 
These were deposited by the ice directly. Their topography is distinctive. 
The surface is characterized by sharp knolls and enclosed basins, by 
hillocks, and hollows, or by interrupted ridges and troughs, which may 
interlock in places and assume rudely parallel positions. The depressions 
are often without outlets and water falling into them forms marshes, 
ponds, and lakes where the material constituting their bottoms is suffi- 
ciently impervious to prevent its escape. The knolls vary in size, but 
are usually rounded at the top and have moderate to steep slopes. Varia- 
tions in height of moraines are due to the unequal amount of material 
held by the ice at those points. The moraines are arranged in rudely 
concentric lines which mark successive positions of the border of each 
ice sheet as it was melted off from that region. For building and foundry 
purposes, it is only the "sandy moraines" that contain deposits of value. 

Ice advanced over Minnesota from two directions. One lobe of the 
glacier came from the north and northeast, from the Lake Superior re- 
gion and northern Wisconsin, traveling over rocks of the siliceous type, 
largely sandstones, quartzites, and crystalline rocks. The material 
brought to Minnesota by this lobe of the continental ice sheet was pre- 
dominantly sandy and such clay as it contained carried relatively small 
amounts of fluxes such as lime, soda, and potash. This material is 
known as the "red drift," considerable deposits of which occur along the 



26 THE FOUNDRY SANDS OF MINNESOTA 

eastern side of the state, from Minneapolis and St. Paul north. By 
careful selection material may be found in this red drift well suited for 
use as sand. Some is used for foundry work at Minneapolis. See Table 
XII, No. 254. 

The other lobe of the glacier coming from' the west and northwest 
brought much more limestone, clay, and shale — the gray drift. The de- 
posits of sand are relatively few and impure. There is little material 
that can be used for foundry work. 

Glacial loams. — Some of the loams in Minnesota are derived from 
moraine material. In creeping slowly over the region the glacial ice. 
1,000 feet or more in thickness, with great quantities of rock incorporated 
in it, developed much movement within the mass itself. It served as a 
great mill in which the rock fragments were ground upon each other, 
much of the rock being reduced to fine silty material which has aptly 
been called "rock flour." During the warmer seasons of the year in 
glacial time, great streams issued from the ice margin carrying large 
quantities of this rock flour which was spread over the flood plains. 
During the colder seasons, with the diminishing of the flood waters, this 
silty rock flour became dry dust which was blown about by the winds 
and some of it was carried back over the glacier's margin and deposited 
on the surface of the ice itself. When the ice ceased to move and finally 
disappeared by melting, this silty rock flour was gradually let down and 
left as a blanket mantling itself indiscriminately over the knolls, hills, 
and intervening low areas of the terminal moraine, as well as over the 
ground moraine. 

This silty rock flour was derived in part from the grinding up of 
partially decomposed feldspars, argillaceous limestones, and siliceous 
dolomites susceptible of further decomposition. Later, under the in- 
fluence of superficial weathering, decomposition and leaching took place 
carrying away much of the soluble matter and leaving behind the more 
insoluble clay residuum, thereby changing the rock flour from a silty 
to a loamy consistency. More or less of the loams in Minnesota that 
are used in foundry work were derived in this manner. These loam de- 
posits are variable in depth, in texture, in clay content, and in content 
of other minerals, and are irregular in distribution. A product uniform 
in character is hardly to be expected from this source, and an attempt 
to map these loams except in a very general way would encounter many 
perplexing problems. 

Outwash deposits. — As the ice melted the water issuing from the edge 
of the glacial ice carried with it sand and gravel. These were spread 
over the plains in the form of alluvial fans. As the escaping waters con- 
tinued to deposit sand and gravel, these fans united and formed outwash 



FORMATIONS YIELDING FOUNDRY MATERIALS 27 

plains. The outwash plains lie on the outer border of the moraines, and 
the sand and gravel composing them is sorted by the action of the escap- 
ing waters. Many outwash deposits may be traced up to a moraine, which 
marks the position of the ice border when they were laid down. Some 
show a decrease in coarseness passing away from the moraine, the coarse 
material having been dropped close to the edge of the ice and the fine 
carried a greater distance. 

The waters found themselves in valleys, over which they spread. 
Carrying loads too great for their velocity, they aggraded their valleys 
and built up large valley trains. At the maximum of their discharge, 
the outwash deposits consisted of sand and gravel. They extend far 
beyond the unstratified drift with which they are associated. They sup- 
ply by far the greater portion of the sands used for structural work in 
Minnesota, and also much foundry material. 

A type of loam differing slightly from that formed in the weathering 
of sandy moraines is derived from the sandy flood plains of glacial 
streams. Probably the broad flood plains supported some vegetation and 
received deposits of sand and rock flour only at time of flood ; so that 
the sand is mixed with vegetation as well as partly weathered. Some 
such deposits of loam are many feet thick. By reason of their mode of 
formation, they are likely to be more uniform in composition and depth 
than the loams of morainic deposits. This type of loam is therefore most 
promising for foundry material. 

The loam on the old flood plains has been subjected to the same process 
of superficial weathering as the moraines ; that is, it has decomposed and 
leached to some extent, as a result of which the upper one to three feet 
is frequently heavier, or more clayey, than that below, which must be 
taken into account in pitting it for foundry use. 

Much of the loam referred to underlies good agricultural land. It 
is possible to recover the loam without destroying the soil. If the upper 
portion of soil is moved by wheel scrapers and spread again over the 
floor of the pit after the foundry loam is removed, the land is still of 
full value for agricultural purposes. Loam can be removed in this way 
at a very low cost without injury to farm lands. A similar method has 
been used in the Red River Valley in working some clay deposits. The 
method is employed in New Jersey to win foundry loams of exactly the 
type here described. 

The sands above described as occurring on glacial flood plains do 
not all border streams at present, for the streams that built these ancient 
plains in many instances became extinct with the glacier. The old flood 
plains are not infrequently far removed from present drainage lines, or 
if they are near present drainage, they may or they may not bear definite 
relation to present drainage. 



THE FOUNDRY SANDS OF MINNESOTA 



Lake beaches. — In the northwest part of the state, in the region of 
glacial Lake Agassiz, there is a shortage of all kinds of sands. The sands 
of this region are secured from ridges of sandy gravel formed along the 
shores of extinct lakes at successively lower levels, down to existing lakes. 
These ridges are known as "lake beaches," and are the result of sorting 
by wave action. 

Loess. — In the southeastern part of Minnesota from the vicinity of 
Red Wing to the Iowa line there is a belt of country bordering the 
Mississippi River about 30 miles wide and 100 miles long, in which a fine 
silty loam known as loess occurs at many places. The distribution of 
the loess is shown in a general way in Figure 2. 



111 1 ■ 1 ■ ■ ■ 1 



20 Miles 




FIGURE 2. MAP OF SOUTHEASTERN MINNESOTA SHOWING AN AREA (DOTTED) IN 

WHICH THE DRIFT IS LARGELY COVERED WITH LOESS. THE! LOESS MANTLE IS 

NEARLY CONTINUOUS EXCEPT ALONG THE LARGER VALLEYS 

AFTER FRANK LEVERETT 

The loess is of interest to the founding industry as a source of mold- 
ing sand for brass, other alloys of copper, aluminum, etc. Most of the brass 
sand imported by the foundries of Minnesota from Missouri, Illinois, 
and Kentucky is simply loess such as may be had in southeastern Minne- 
sota at hundreds of places. 



FORMATIONS YIELDING FOUNDRY MATERIALS 29 

The loess occurs in the area shown in Figure 2 up to 40 feet in 
depth, and possibly deeper. It is not limited by any altitudes in this 
area, but occurs capping the highest hills, bluffs, ridges, and uplands 500 
feet or more above the Mississippi, as well as in the valleys down prac- 
tically to the river level. A typical example of loess is to be seen at Red 
Wing capping Barn's Bluff which is a small isolated hill rising with 
~brupt slopes, and sheer cliffs to about 400 feet above the river. The 
Dess, at least 20 feet in depth, caps the highest part of this hill. It 
xcurs also in the vicinity of Red Wing mantling the terraces 40 feet 
above the Mississippi and back from the main stream in the side valleys 
at all levels to points 400 feet above the river. 

Some of the loess at low levels has been derived from the upland 
loess and washed down by streams, but part of it was originally deposited 
at the lower elevations. 

On the whole the loess on the uplands 200 to 500 feet above the 
Mississippi is more uniform in texture and mineral composition than the 
loess in the valleys. In deep fresh exposures the loess is usually pale 
blue-gray in the lower portion, pale yellowish gray in the middle portion, 
and brown or yellowish brown in the upper one to three feet. There is 
an appreciable difference usually in the texture of the material accom- 
panying the difference in coloring; that is, the basal portion of bluish 
tint is more silty, the central portion more loamy, and the surface por- 
tion clayey. The difference in coloring and texture appears to be due 
to progressive weathering rather than to a difference in the original 
deposition. 

Owing to the great abundance of the material and its accessibility 
to the railroads it is unnecessary to describe in detail the numerous ex- 
posures. At Clay Bank station, south of Red Wing, the Chicago Great 
Western Railroad has made a cut about 8 feet deep through the loess 
and the material could be loaded there with steam shovel. Other deposits 
of the same material similarly situated with reference to railroads could 
no doubt be found so that for foundry needs a practically unlimited sup- 
ply can be had at nominal cost. 

The following analysis shows that the loess carries about 13 per cent 
of iron, alkali, and alkali earth compounds, which would insure a low 
fusion point, but since the temperatures of the melts of brass and its 
alloys are low there is no danger of fusion of the loess when used as 
molding sand for these metals. 



30 THE FOUNDRY SANDS OF MINNESOTA 

The texture of the loess is indicated by the mechanical analyses of 
Table XL A chemical analysis is stated below : 

CHEMICAL ANALYSIS OF LOESS (BY F. F. GROUT) 

Silica 71-53 

Alumina 8.07 

Iron oxide 5-63 

Magnesia i-74 

Lime 2.36 

Soda 1 .85 

Potash 1-97 

Ignition 4-50 

Water 2.30 

Titanium oxide -3 1 



100.26 

As to the origin of the loess there has been much difference of opinion, 
some investigators claiming a wind origin and others a water origin. 
Probably both agencies have been active. Undoubted evidence of water 
stratification in the loess is to be seen in certain localities even on the 
uplands 500 feet above the Mississippi River ; and at other localities 
equally conclusive evidence of wind work is to be found. There has, no 
doubt, been more or less redistribution of the loess by both wind and 
water since its original deposition. 

The glacial rock flour, above described, may have been the chief source 
of the loess, but the fact that the impure dolomites of the region upon 
weathering leave a residuary product of closely similar constitution sug- 
gests that this also may have been a source. 

Recent deposits. — The sands of the recent period used in Minnesota 
belong to two classes, lake sand and river sand. Most of the lake sand 
can be used for building purposes, and may be used in some foundry 
work. An extensive plant is operating at Duluth where the sand is ob- 
tained from Lake Superior. 

A type of loam is found mantling the terraces in the valleys of streams- 
at levels of 20 to 200 feet above present streams. In mode of origin 
these loams are not unlike the glacial outwash loams. In fact some of 
them are genetically the same. Others, however, were formed in a 
similar manner but at a later peroid ; that is, they were deposited by the 
streams that in postglacial time have passed through the stages of ag- 
gradation and degradation. Generally these loams are variable in char- 
acter, but afford good foundry material if sufficient care is exercised 
in choosing the localities and in pitting the material. 



FORMATIONS YIELDING FOUNDRY MATERIALS 



3i 



TABLE IV. LIST OF MINNESOTA FOUNDRIES AND LOCALITIES SUPPLYING 

THEM WITH FOUNDRY SAND 



Brass Sand 



Molding Sand 



Core Sand 



Albert Lea 

American Gas Machine Co. 
Austin 

Austin Foundry 

Brainerd 

Parker and Topping Co.... 



Chaska 

Ess Bros 

Cloquet' 

Cloquet Foundry 

Crookston 

Crookston Iron Works. 
Duluth 

Clyde Iron Works... 



Duluth Brass Works 



Duluth Foundry and Faucet Co. 
Duluth Iron Works 



Minnesota Radiator Co. . . . 

National Iron Company.. 
Fairmont 

Fairmont Gas Engine Co. 
Faribault 

Faribault Machine Shop . . . 

Winter and Co. . 

Nutting Truck Co 

Fergus Falls 

Fergus Falls Iron Works. 
Hibbing 

Oliver Mining Co 



La Crescent 

Smith Grubber Co 

Lake City 

Gillett-Eaton and Squire Co. 
Mankato 

Little Giant Co 

Mankato Mfg. Co 

New Prague 

New Prague Foundry Co.... 
Ortonville 

Ortonville Foundry 



Owatonna 

New Owatonna Mfg. Co. . 

North Star Iron Works. 
Paynesville 

C. W. Peavey 

Red Wing 

Red Wing Iron Works.. 
St. Cloud 

Granite City Iron Works. 

St. Cloud Iron Works... 



Local 



Chicago 



Outside state 

Lime Springs, la. 

Kerrick 
St. Paul 

Minneapolis 
Local 
Kerrick 
Kerrick 



Whitehead Bros., 

Buffalo, N. Y. 
Albany sand 



St. Peter, Minneapolis 

Local 

Coarse, local 
Fine, St. Peter, 
Minneapolis 

Minneapolis 

Local 

Local, glacial 

Fine, St. Peter, 

St. Paul 
Coarse, Lake Superior 
Whitehead Bros., 

Buffalo, N. Y. 



Kerrick 


Coarse, Lake Superior 




Fine, St. Peter, 




St. Paul 


St. Paul 


Lake Superior 


Kerrick 


St. Peter, St. Paul 


Waterloo, la. 


St. Paul 


Local 


St. Peter, local 


Local 


St. Peter, local 


Local 


St. Peter, local 


St. Paul, Kerrick 


Local, glacial 


St. Paul 


Local, glacial 


Local 


Local 


Local 


Local 


Local 


St. Peter, local 


Local 


St. Peter, local 


Mankato 


Mankato, St. Peter 


Big Stone City, 


Local, glacial 


S.D. 




Faribault 


St. Peter, Faribault' 


St. Paul 


St. Peter, Faribault 


Local 


Local, glacial 


Local 


Local 


Minneapolis 


St. Peter, St. Paul 


Kerrick 


St. Peter, St. Paul 



32 



THE FOUNDRY SANDS OF MINNESOTA 

TABLE IV— Continued 



Brass Sand Molding Sand 



Core Sand 



Thief River Falls 

Thief River Falls Iron Works 

Virginia 

Virginia Foundry Co 

Winona 

Gate City Iron Works 

New Winona Mfg. Co 

Winona Machinery and Foundry Co 
Minneapolis 

American Brake Shoe Co 

Crown Iron Works 

Commutator Co 

Eagle Foundry 

Flour City Ornamental Iron Co 

Gas Traction Foundry 

Mpls. Steel and Mach. Co 

Soo Line Railroad Shops 

University of Minnesota Foundry. . . 
St. Paul 

American Hoist and Derrick Foundry. 

Herzog Foundry 

Northern Malleable Iron Works.... 

Union Brass Works 

Valley Iron Works 

St. Paul Foundry Co 



Albany, N. Y. 

Milwaukee, Wis. 

France 

St'. Louis, Mo. 

Albany, N. Y. 
Albany, N. Y. 

Albany, N. Y. 

Local, glacial 
St'. Louis, Mo. 
Fort Snelling 
Monmouth, 111. 



St. Paul, Kerrick 
Kerrick 

LaCrosse, Wis. 

( Rockford, 111. 

1 Minnesota City 
Minneapolis 

Local, glacial 
Local, glacial 

Local, glacial 

Local, glacial 
Local, glacial 

Local, glacial 

Local, glacial 



Local, glacial 
Local 



Local 

Biwabic, glacial 

Minneapolis 
River, local 
Jordan, Ottawa 



f St. Peter, local 
1 Glacial, local 
St. Peter, local 

Steel, Jordan, Ottawa 

St. Peter, local 
St. Peter, local 

Steel, Ottawa 
St. Peter, local 



LABORATORY METHODS OF TESTING FOUNDRY SANDS, 

LOAMS, AND CLAYS 

Mechanical analysis of sands, loams, and clays. — The mechanical 
analysis of a sand or loam, as generally understood, consists simply in 
separating the material into a series of products graded according to size 
of the grains and weighing these products to determine the percentage 
of the coarse and the fine material present. There are many ways of 
sizing, but most of those in general use are modifications of two well- 
known methods : ( i ) by using a set of screens, in which the material 
is separated dry into a series of products ranging from coarse to fine, 
and (2) by the use of water so that the different sized grains in the 
material are separated from each other by their buoyancy. The former 
is commonly called the dry method, and the latter the wet method. The 
two methods are frequently combined and there is almost no limit to the 
degree of refinement to which either method may be carried. 

The sizing of sands by means of a set of screen sieves is employed 
universally where sands are used in the various industries and the value 
of a given sand, or its adaptability for a certain purpose, is based to a 
greater or lesser degree on the results shown by the screen analysis. 

The dry screen analyses of the molding sands were made in the 
ordinary way, using a set of standard 6-inch screen sieves of 4, 10, 20, 



LABORATORY METHODS OF TESTING SANDS 



33 



30, 40, 50, 60, 80, 100, and 200 meshes to the inch. The 30- and 50-mesh 
screens were used only in special cases. The sand samples for analysis, 
consisting of 50 grams (about 2 ounces), were previously dried in a 
sand bath at a temperature of ioo° C. for 24 to 48 hours, or until they 
eased to lose weight, after which the samples were pulverized in an iron 
lortar with a rubber shod pestle to break up the clusters without break- 
ing the individual sand grains. The screens were shaken by hand, using, 
one, two, or more screens at a time, as conditions warranted. In some 
instances a small stiff bristle brush was used on the screens to hasten 
the passage of the sand through the screen, since with some samples no 
reasonable amount of shaking or jarring would pass the sand without 
the aid of a brush. A mechanical shaker operated by an electric motor 
was tried, but found impracticable, since uniform or complete separations 
generally could not be obtained. 

The screens were calibrated in the physics laboratory of the Univer- 
sity of Minnesota by the aid of an optical micrometer, with the results 
>hown in Table V. 

TABLE V. SIZE OF SCREEN OPENINGS AND DIAMETER OF WIRE IN SCREENS 





Diameter in Inches 


Size in Inches of 


Size in Inches of 




of Wire in Screens 


Opening of Screens 


Opening of Screens 


Mesh of 


According to 


According to 


Found by Actual 


Screen 


Manufacturer 


Manufacturer 


Measurement 


4 


.065 


.185 


.185 


10 


.03s 


.065 


.074 


20 


.0172 


.0328 


.0341 


30 


• 0135 


.0198 




40 


.01 


.015 


.0189 


60 


.008 


.0087 


.0096 


80 


.0057 


.0068 


.0074 


100 


.0045 


.0055 


.0062 


200 


.0021 


.0029 


.0038 


220 


.0017 


.0028 




240 


.0016 


.0026 




260 


.0016 


.0022 




280 


.0016 


.002 




300 


.0016 


.0017 





In ordinary sands the grains are more or less coated by a film oi 
clay which in the fine sizes materially increases the actual diameter of 
the grains. In sands where silty material is present in appreciable amounts 
together with 2 to 10 per cent of clay, the silt grains are cemented to the 
coarser grains of sand by the clay film in greater or less amounts which 
still further increases the actual diameter of the sand grains. Both the 
silts and the finer sizes of sand are gathered in clusters, or compound 
grains, particularly in loamy sands, and these clusters function as unit 
masses, and in the ordinary screen analysis appear as individual grains 
in the various sizes. 



34 



THE FOUNDRY SANDS OF MINNESOTA 



The purpose of pestling dry sand is to break up these clusters, free 
the sand grains from adhering finer material, and bring the material 
into a state of individual particles. In some sands, and in loams gen- 
erally, however, no ordinary amount of pestling in the dry state will 
completely break up the clusters or remove the adhering silt grains, much 
less will pestling remove the clay coating from the grains of sand. Hence 
it is impossible to obtain a true sizing of the material by the dry method. 
The common screen analysis, while valuable because easily made and be- 
cause it shows in a general way the physical constitution or texture of 
the sand, is at best a proximate analysis only. 

Inasmuch as clay in molding sand performs an important function as 
bonding material, it is essential that a determination as accurate as possible 
be made of the clay present in the different molding sands, and accord- 
ingly the following procedure, adapted from one of the common methods 
of soil analysis, was employed. 

A 50-gram sample of the sand to be tested was taken in the same 
manner as for a dry screen analysis ; that is, the sample was weighed out 
after thorough drying in the sand bath. The sample was transferred to a 
common 8-ounce nursing bottle. Six ounces of distilled water and 5 cubic 
centimeters (about one teaspoonful) of ammonia, were added, making the 
bottle about half full. The bottle was placed horizontally in a mechanical 
shaker, shown in Figure 3, and shaken for five hours ; the purpose being 
to free the sand grains of any coating of clay and to deflocculate the clay. 




FIGURE 3. HOT AIR ENGINE USED FOR SHAKING BOTTLES IN CLAY 

DETERMINATIONS 



LABORATORY METHODS OF TESTING SANDS 35 

The shaker was propelled by a small hot air engine. The cylinder C 
about 2 inches in diameter by 12 inches long, was thin walled. The upper 
part was surrounded by a tank of water W to prevent excessive heating. 
Heat was supplied at the lower end by a Bunsen burner A, the amount of 
heat being controlled by the supply of gas, this in turn determining the 
speed of the engine. The engine would run for hours without attention. 
The length of stroke of the engine piston was about 2 inches. The tray T 
carried 12 bottles and moved horizontally on the track R. The flywheel 
F ran about 40 revolutions per minute. The length of stroke of the pit- 
man from the flywheel to the tray was about 3 inches. 

Immediately after shaking, the contents were washed into graduated 
glass cylinders about 13 inches tall and ij4, inches in diameter which were 
allowed to stand at room temperature for 24 hours. The material that 
remained in suspension after 24 hours was considered clay. The water 
with clay in suspension was siphoned off into large beakers, after which 
the sand in the cylinder was again agitated by jetting distilled water under 
pressure into the sand ; then the cylinders were allowed to stand again for 
24 hours. This process was repeated until the cylinders showed nothing 
in suspension on standing for 24 hours. 

The material obtained from these 24-hour decantations, designated clay, 
was precipitated in the beakers by adding a small amount of barium 
chloride, for in most cases the clay would not settle out completely by 
gravity. This clay was recovered, dried, and weighed, thus giving the per- 
centage of clay in the sample which might be regarded as acting as the 
bond for molding purposes. 

The contents of the cylinders, after removing the clay, were again 
agitated with distilled water and allowed to stand for 30 minutes. The 
water, with what was in suspension, was siphoned off, or decanted, at the 
end of 30 minutes, this process being repeated until nothing remained in 
suspension for the 30-minute period. The material from these 30-minute 
decantations, called "thirty minute silt," was recovered, dried, and weighed. 

In the same manner a 15-minute silt was separated, recovered, dried, 
and weighed, after which the contents of the cylinders, regarded as true 
sand, were recovered, dried, and weighed, and then subjected to the usual 
mechanical dry screen analysis, using the screen sieves. 

This combination of wet and dry methods of analysis shows the per- 
centages of clay, 30-minute silt and 15-minute silt, present in the sample 
and the percentage of 4, 10, 20, 40, 60, 80, 100, and 200-mesh sand present 
after removing the clay and silts, making 11 separates in all. 

In the course of the laboratory work on foundry sands, 320 samples 
of sand, loam, and clay were subjected to the common screen analysis, fol- 
lowing the dry method above described. Two hundred sixteen of these 



36 



THE FOUNDRY SANDS OF MINNESOTA 



samples were subjected to the combined wet and dry method of analysis. 
In the various tables where the results of these analyses are recorded, the 
dry method analyses are referred to as "A" and the analyses by the com- 
bined method, showing the percentage of clay and silt, are referred to as 
analyses "B." 

Comparison of wet and dry methods of analysis. — A comparison of 
the results of these two methods of analysis of the same materials brings 
out many facts in connection with the texture of sands. Illustrating the 
marked difference in results by the two methods, the results of analyses 
of three samples, 9, 3, and 18 are given in which analysis "A" and "B" in 
each are duplicate samples of the same material. These examples are 
fairly representative of the entire series. 



TABLE VI 



Mesh of Screen 



4 — 10 .... 

10 — 20 .... 

20 — 40 .... 

40 — 60 .... 

60— 80 

80 — 100 .... 
100 — 200 .... 
200-silt 

15-minute silt 

30-minute silt 
Clay 

Total . . . 



Sand 


No. 9 


A 


B 


.80 


• 74 


43-50 


22.80 


29.00 


26.06 


17.00 


25.08 


4-50 


10.00 


1. 00 


7-44 


2.00 


3-50 




.12 




1.40 




2.00 


97.80 


99-14 



Sand 


No. 3 


A 


B 


.20 




.82 


.86 


6.63 


5-74 


18.34 


14.10 


6.20 


6.08 


4.64 


4.22 


16.90 


10.98 


43-32 


38.00 




2.00 




6.00 




10.80 


97-05 


98.78 



Sand 


No. 18 


A 


B 


1. 00 


.20 


.20 


• 30 


.70 


.60 


1. 00 


• 54 


1.50 


2.32 


6.00 


3.38 


29.70 


22.82 


59-50 


53-48 




1.90 




7.00 




7.00 


99.60 


99-54 



Sample 9 is a quartz fire sand, 3 is a common loam, and 18 is a brass 
sand. Notwithstanding all the samples for analyses "A" were thoroughly 
pestled, there still remained clusters or aggregates of grains that did not 
break up into individual particles. These clusters did break up by treat- 
ment with ammonia in the analyses "B." 

It is to be noted that different sands and loams behave very differently 
with respect to the clustered grains ; that is, some samples yield to pestling 
readily while others do not, so that there is no uniformity in results, thus 
the sizing obtained by the dry method of analysis is not to be depended 
on even in sands with a small clay content, as sample 9 shows. 

The primary purpose in the wet method analysis was to determine 
the percentage of clay present in the various sands, loams, and so-called 
clays used in foundry work, the assumption being that such clay afforded 
the bond essential to molding sand. But this method of analysis also pro- 
vides two silt determinations; that is the 15-minute and 30-minute silt, the 



LABORATORY METHODS OF TESTING SANDS 37 

grains of which are intermediate in size between the 200-mesh screen and 
the clay. Such experiments as were undertaken indicate that these silts 
play an important role in the permeability of the sand by their behavior in 
granulation. They are important also in connection with the tensile 
trength developed in the molds, both in the green and in the dry sand. 

Perfect sizing was not attained in the wet analysis "B" because the 
method employed involved two essentially different principles in the 
separation. That is, the silt separations relied on the suspension of the 
particles in water, which in turn depended on the specific gravity of the 
individual particles and not on their mass or diameter, whereas the sizing 
of the sand after the silts and clays were removed, using the sieves, de- 
pended on the diameter of the grains only. 

As a check on the method of sizing, the silts obtained in the analyses 
"B" were examined under a microscope, using a scale to measure the 
diameter of the silt grains. While, as was anticipated, many large grains 
were found in the silts, some even larger than the openings in the 200-mesh 
screen, still the percentage of such grains was small, and the bulk of the 
material, both in the 15- and the 30-minute silt, was fairly uniform in size. 
These measurements showed that the average diameter of the grains in 
the 15-minute silt was .00053 inch and the average diameter of the grains 
in the 30-minute silt was .000236 inch. That is, the diameter of the 30- 
minute silt grains was a little less than half that of the 15-minute silt, indi- 
cating a fair degree of sizing by water suspension. 

The openings in the 200-mesh screen, as shown in Table V, were .0038 
inch ; and while no attempt was made to measure, or otherwise determine 
the size of the clay grains, numerous workers in soil analysis have given 
the diameter of clay grains as "under .000197 inch." From which it will 
be seen that there is a rather wide interval between 200-mesh sand and 
clay, of which heretofore no account has been taken in ordinary molding 
sand analyses. The determination of the percentage of 30-minute and 
15-minute silts present in the various materials makes possible a more 
critical comparison of the samples as to their texture and permeability. 

Comparing analyses. — To facilitate comparison of analyses a factor 
was sought that would express in a general way the comparative fineness 
of the individual sample as indicated by the screen analyses. Such a 
factor is recorded in the tabulated analyses. 

The diameter of the sand grains in any of the separates obtained by 
screening naturally varies from the size of the opening of the screen 
through which the sand passes to the size of the opening of the next finer 
screen on which the sand is caught. The average grain in each of the 
separates may be taken as the mean between the openings. For example, 
the openings in the 40-mesh screen were found to be .0189 inch and the 



38 



THE FOUNDRY SANDS OF MINNESOTA 



openings in the next finer screen, viz., 6o-mesh, .0096 inch. The mean be- 
tween these two is .01425 inch and the average or mean sized grain of the 
40- to 60-mesh sized sand is taken to be .01425 inch in diameter. Determin- 
ing the mean or average diameter of the other separates in the same man- 
ner gave the results shown in the following table. 

TABLE VII 



Mesh of Screen Used in 


Average Diameter of Sand 


Sizing Sands 


Grains in Sized Sands 


4 — 10 


.1295 inch 


10 — 20 


.054 


20 — 40 


.0265 


40 — 60 


.01425 


60 — 80 


.0085 


80 — 100 


.0068 


100 — 200 


.0050 


200-silt 


.00216 


15-minute silt 


.00053 


30-minute silt 


.000236 


Clay — below 


.000197 



The diameters of the 15- and 30-minute silts were determined by 
measuring the silt grains themselves with a microscope. The diameter of 
the clay grains was taken from tables published by workers in soil analysis. 

The method of computing the fineness factor, or comparative grain, 
was to multiply the average diameter of the grain in each separate by the 
percentage of the separate present in the sample and divide the sum of the 
products so obtained by the sum of the percentages of the several separates. 

For example, brass sand 18, taken from Table XI, computed for grain 
and mesh, gives the following results : 

TABLE VIII 

Average Diameter Product of Per Cent 

of Sand Grain Proportion of Each Multiplied by Average 

Mesh of Screen in Each Separate Separate Present Diameter of Grain 

4 — 10 -1295 X -002. r= .000259 

10 — 20 .054 X .003 = .000162 

20 — 40 .0265 X .006 = .000159 

40 — 60 .01425 X -0054 = .000076 

60 — 80 .0085 X .0232 = .000197 

80 — 100 .0068 X -0338 = .000229 

100 — 200 .0050 X .2282 = .001141 

200-silt .002165 X .5348 = .001157 

15-minute silt .00053 X .019 = .000010 

30-minute silt .000236 X .07 = .000016 

Clay .000197 X .07 = .000014 



• 9954 



.003420 



.OO.342O 

■9954 



= .0034 inch, the comparative diameter of grain. The com- 



parative mesh computed by this method is an approximation only. 
Greater accuracy would have been obtained if the number of grains of 



LABORATORY METHODS OF TESTING SANDS 



39 



sand in a unit volume for the different separates had been computed and 
used as a factor in computing the comparative mesh. 

The comparative mesh was obtained by interpolating the comparative 
diameter of grain in the scale of screen openings given in Table V (page 
33). Interpolating .0034 inch in the scale of screen openings, it was found 
to correspond to a mesh of 216 (see Fig. 13). Accordingly 216 is taken 
as the comparative mesh of this sand. 

Laboratory tests for bonding power of sands, loams, and clays. — 
In testing the bonding power of molding sands the usual method em- 
ployed in testing cement for tensile strength was used. This consists in 
making a briquet of the material to be tested having a cross section of one 
square inch in the center and larger at the ends, as illustrated in Figure 4, 
A, B, and C. The briquets were broken in the apparatus, Figure 5, by 





FIGURE 4. BRASS MOLD FOR 
MOLDING SAND BRIQUETS 
FOR TENSILE STRENGTH 
TESTS 



FIGURE 5. APPARATUS FOR BREAKING 
BRIQUETS OF MOLDING SAND TO DE- 
TERMINE TENSILE STRENGTH 



applying a load at the lower end of the briquet and measuring the stress, 
in pounds' or ounces, required to break the one-square-inch section. The 
results give the tensile strength of the molding sand in pounds per square 
inch. In making the briquets the molds, shown in Figure 4, were used and 
the molding sand was rammed into them, following the approved methods 



4 o THE FOUNDRY SANDS OF MINNESOTA 

of foundry practice in tempering the sand and tamping it into the mold. 

The briquets were run in series of three, designated series A, B, and C, 
with three briquets in each series. The three briquets in series A were 
broken at once upon taking them from the molds to determine the tensile 
strength of the damp or green sand, taking the average of the three tests. 
The three briquets in series B were allowed to stand in the open air 24 to 
48 hours for room drying, after which they were broken to determine the 
tensile strength of sand comparable to that of air-dried molds in foundry 
practice. The three briquets in series C were kiln dried, or furnace burned, 
as circumstances might suggest, at temperatures ranging from 400 C. 
(752 F.) to 1000 C. (1832 F.) after which they were broken and 
tensile strength noted. 

In view of the fact that green molding sand rarely develops a tensile 
strength of more than a few ounces to the square inch, whereas the ma- 
chines for testing cement briquets are not sensitive to such light loads, it 
was obvious that these machines could not be used for breaking the mold- 
ing sand briquets, so special apparatus had to be devised for that purpose. 
A pair of jaws (Figure 5 T and B) were made of mahogany and provided 
with a mounting so that the briquets could be suspended vertically and 
the load applied directly to the lower end. This apparatus was sensitive 
to a breaking load of 2 ounces and gave satisfactory results up to loads 
of 100 pounds. 

The upper jaw was suspended at H, by means of a hook and eye, and 
the lower jaw was carried by the briquet itself ; thus forming a couplet 
that swung free and eliminated all torsional stresses. A bucket L was 
suspended on the hook / to receive the load, which consisted of water that 
was run into the bucket L by means of a rubber tube and pinchcock. 

In measuring the breaking load everything below the breakline was 
counted ; that is, the bucket L with its contents, the lower jaw B , and the 
lower half of the briquet carried by the jaw B were weighed and counted 
as breaking load. The blocks SS served to catch the jaw B when the 
break occurred so that there was no loss by spilling the water. 

Green briquets weighed from 95 to 120 grams, or about 3 to 4 ounces," 
and in this method of testing the strength, about half of the weight of the 
briquet itself, was counted as initial load, and this was the measure of the 
sensitiveness of the apparatus ; that is, if the briquet broke of its own 
weight its breaking strength was figured as one half of the weight of the 
briquet itself, or something less than two ounces. 

For heavier loads larger buckets were used in the place of bucket L 
and bird shot, such as is used in ordinary cement briquet-breaking ma- 
chines, was employed as load. 



LABORATORY METHODS OF TESTING SANDS 41 

Determining the porosity of foundry sands and loams. — By porosity 
of a sand is meant the voids or interstitial space between the sand grains 
or the granules. It is the space unoccupied by solid matter, but occupied 
by air or water. The porosity of a sand is usually expressed in percent- 
ages of the volume of the mass. If a sand has a porosity of 30 per cent, 
it means that the sand grains occupy only 70 per cent of the actual volume 
of the container, and the remaining 30 per cent is vacant space or is 
occupied by air. 

The test for porosity of the sands consisted in comparing the weight 
of a standard volume of each sample with the weight of the same volume 
of solid non-porous rock. 

The standard volume used in the porosity tests was a brass cylinder 
3 inches deep by 2 13/32 inches in, diameter (Fig. 6 V), which was ob- 
tained by cutting a section from standard rolled brass tubing. This volume 
of water at ordinary room temperature weighed 228 grams. The average 
specific gravity of the rocks of the earth is 2.6 to 2.7. Ordinary molding 
sand is simply the fragments and weathered products of such rocks, and 
the numerous determinations made by many independent workers in soils 
and cement, and investigators of rock-weathering, show that common sand 
has a specific gravity essentially the same as average rock, viz., 2.65. The 
variations in specific gravity for the different types of sand are so slight 
as to be negligible. The specific gravity of the "fines" is slightly higher 
than for the coarse constituents, so that we have assumed a specific gravity 
of 2. J as probably as near the truth as it would be possible to get even if 
precise determinations had been made. Multiplying 228, the weight of 
the standard volume of water, by 2.7 gives 615.6 grams as the weight of 
this standard volume of non-porous rock. 

The standard volume brass cylinder was filled in turn with each of the 
sands after they had been dried in the sand bath for 24 hours, and a 
standard method of jarring the cylinder was used to compact the sand, 
following the method described by King. 6 

In making 1 the porosity tests on molding sand the method of King was 
modified slightly as follows: A section of brass tubing (Fig. 6 D) the 
same diameter as the cup (Fig. 6 V) and 1^2 inches high, was placed on 
top of the cup, held in position by a band. This section was filled with 
sand, thus adding a head of 1^ inches to the receptacle. It was found 
that more uniform results could be obtained in compacting the sand by 
this means. 

6 King, F. H. Principles and conditions of the movements of ground water: Nineteenth 
Annual Rept. U.S. Geol. Survey, 2, p. 208, 1899. Some of the preliminary laboratory work in 
determining the porosity and permeability of the sands discussed in the treatise above mentioned 
was done by the writer in 1897 under the direction of Professor King. 



42 



THE FOUNDRY SANDS OF MINNESOTA 



The method of compacting the loose sand, as described by King, con- 
sists in holding the cup filled with sand firmly on a rigid table, preferably 
a stone slab, with one hand, and jarring the cup with the other hand by 
tapping it with a piece of wood, turning the cup part way around from 
time to time ; the tapping of the cup being vigorous enough to disturb the 
sand grains and allow them to rearrange themselves. 




FIGURE 6. BRASS CYLINDERS USED FOR MOUNTING CORES OF MOLDING SAND FOR 
PERMEABILITY TESTS J AND CUP V USED IN POROSITY DETERMINATIONS 



LABORATORY METHODS OF TESTING SANDS 43 

After the sand had settled as much as it would by jarring, the extra 
section was removed from the cup, the top of the cup was stricken off 
with a metal straight edge, and the cup and contents weighed. Subtracting 
the net weight of the sand in the cup from 615.6, the weight of the same 
volume of solid rock, and dividing the difference by 615.6 gives the pore 
space in the sand in percentage of volume. 

Other methods of determining the pore space were tried, such as add- 
ing sand to a measured volume of water and noting the volume of water 
displaced by the sand, but the results were unsatisfactory because of en- 
trained air. 

More refined methods of determining the pore space did not seem war- 
ranted because the per cent of pore space in sand is not a dependable 
measure of the permeability. Even in loose clean sand the permeability 
can be judged only in a very general way from the porosity, and in damp 
loams where granulation takes place there is no necessary relation between 
the actual porosity and the observed permeability. Permeability on the 
other hand is of vital importance, and this was determined directly by 
testing the sand without reference to its porosity. 

The porosity of a large number of dry sands was determined by the 
method above described ; that of the damp sands and cores was computed 
from the known weight, volume, and other factors ; and the results appear 
in the various tables throughout the report. The porosity determinations 
included in the tabulations show that the permeability commonly is inde- 
pendent of the porosity. 

Testing the permeability of molding sands. — Permeability of sand is 
the facility the sand offers to the passage of a liquid or gas through it. 
Permeability in sand is usually measured by the time required to pa:s a 
given volume of air through a column of sand of known dimensions. 

For testing permeability, apparatus was designed and built that would 
deliver 74 liters of air at a constant pressure of 50 inches of water. In 
some respects this apparatus was adapted from the aspirator used by 
King. 7 A standard column of sand 2 13/32 inches in diameter and 3 
inches high was used in all permeability tests. The time required to pass 
the 74 liters of air through the columns of sand was measured by a stop 
watch, which could be read to a fraction of a second. 

One hundred eighty samples of sand and loam were tested for perme- 
ability with this apparatus, and an average of six runs was made on each 
sample. The time of passage of the air ranged from 24 seconds to 4,646 
seconds in the different samples. The pressure of the air remained prac- 
tically constant, varying less than J / 2 of one per cent during the rapid runs. 
and being absolutely constant during the slower runs. 

7 Op. cit., p. 22.3. 



44 



THE FOUNDRY SANDS OF MINNESOTA 



The apparatus used to test the permeability of sands is shown in 
Figure 7. The apparatus consists of a series of chambers made of 
large threaded iron pipe in which the air was compressed by admitting 
water into the chambers. The apparatus is connected directly with the 
city water system, which has a pressure of 65 pounds, thus insuring an 
ample supply of water as rapidly as may be needed. The water is ad- 
mitted into the apparatus through a 2-inch gate valve at /. It passes 




FIGURE 7. APPARATUS USED IN PERMEABILITY TESTS DELIVERING 74 
LITERS OF AIR UNDER CONSTANT PRESSURE OF 50 INCHES OF WATER 



through chambers EA and AB and into CD, each of which is 6 inches 
in diameter, compressing the air as it advances. A weir 6 inches across 
is provided at L, over which the water passes into chamber CD. A stand 
pipe 4 inches in diameter is provided at F, open at the top, with an over- 
flow for waste at G. The waste gate G is 50 inches above the weir L, 



LABORATORY METHODS OF TESTING SANDS 45 

which insures a uniform head of 50 inches of water, giving a pressure 
of about 4.5 pounds. 

The compressed air is conducted through a 2-inch pipe Q, which ex- 
tends 20 feet above the apparatus, returning to R, where wet and dry 
bulb thermometers are installed in the air line. From R the air is con- 
ducted downward through a common well cylinder in which the sand 
sample to be tested is mounted at vS\ 

After passing through the sand ^ the air is carried upward and 
through a waste valve at T in which another set of wet and dry bulb 
thermometers is installed. A water manometer OPP is connected with 
the air line Q and provided with a scale, graduated to 1/10 of an inch. 
The manometer registers the pressure, 50 inches, direct, and variations 
in pressure can be read to 1/10 inch, or .2 of 1 per cent of the total pressure. 

Chambers AB and CD are provided with glass water gauges MM 
and NN, which show the height of the water in these chambers at all 
times. The volume of chamber CD was determined by measuring water 
into it when the apparatus was installed. The glass gauge MM was pro- 
vided with a scale, and the calibrations in half liters were marked on this 
scale as the water was measured into the chamber CD. The chamber 
CD holds 74 liters to the level of the weir h, which is the capacity of 
the apparatus. When in operation the amount of air that is passed 
through the sample is registered at all times by the position of the water 
level in gauge MM. 

Waste valves are provided at K and / to empty the apparatus of 
water after a run is completed. A baffle at X serves to make the greater 
part of the air in chamber AE available for compression, and to prevent 
undue agitation in chamber AE when the city water under high pressure 
is admitted. 

Control valves are provided at U, V , and W . The wet bulb ther- 
mometers at T and R have the bulbs wrapped with lamp wicking, and a 
small chamber filled with water is provided below the air line beneath 
the wet bulbs, in which the wicking is submerged, thus providing the 
wet bulbs with water to meet the needs of evaporation. 

The upper end of the waste pipe Y is connected with the upper end 
of the manometer P to guard against, and register any back pressure that 
might result from friction or otherwise when air is passed rapidly. 

The procedure in testing is as follows: The sand to be tested is 
mounted in the well cylinder at S; the valves U and V are closed; the 
valve / is opened, admitting the water into the apparatus rapidly so that 
it overflows in considerable volume at G. The rise of the water in 
chamber AB is registered in the glass gauge NN, and the rate the water 
rises here naturally declines as the pressure increases, The manometer 



46 THE FOUNDRY SANDS OF MINNESOTA 

OPP records the pressure as it progresses. When the water in cham- 
ber AB reaches the level of the weir L, which is indicated by a calibration 
mark on the gauge NN, the manometer is read. The valve U is then 
opened allowing the compressed air to pass through the sample. When 
the water in chamber CD reaches the zero mark on gauge MM the stop 
watch is started and allowed to run until the water in chamber CD, as 
registered on the gauge MM, reaches the calibration 74, indicating that 
74 liters of air have passed through the sand. The watch is stopped the 
instant the gauge MM registers 74, so that the time is recorded to a frac- 
tion of a second. The valve / is then closed, and the valves / and K 
opened to empty the apparatus. 

The temperature of the room and the barometer is recorded at the 
beginning of each run, and the wet and dry bulb thermometers at R and 
at T are read three times or more during each run, viz., the beginning, 
at the middle, and at the end of the run. 

The apparatus was readily tested against any possible leaks from de- 
fective mounting or improperly seated valves by allowing a little water I 
to come over the weir into chamber CD, then closing the valve U, and 
allowing the water to run, wasting at the overflow G. If there were any 
air leaks the water would rise in the chamber CD and register on the 
gauge MM. Testing the apparatus consisted in allowing it to stand for 
a period equal to the time required to pass the 74 liters of air through 
the sample; that is, if it required 100 seconds to pass 74 liters of air 
through the sample, the apparatus was allowed to stand 100 seconds 
under full pressure. If the leaks were not appreciable in that time, they 
were neglected. 

The pressure rarely varied more than .3 of an inch during a single 
run, which is .6 of 1 per cent. The variations in pressure were greatest 
with rapidly passing air ; that is, when the time of passing the air was 
50 seconds or less, the variation in pressure might be as high as .6 of 1 
per cent, but when the time was 200 seconds or more, the variation in 
pressure was rarely more than .1 inch on the manometer, or .2 of 1 per 
cent. The average pressure for hundreds of tests was 50.45 inches. 

The apparatus used for mounting the samples of sands is somewhat 
elaborate, due to the wide range of material tested. The main con- 
tainer consists of a common, brass lined, well cylinder 3^ inches in 
diameter and 12 inches long, the plunger and lower valve being removed, 
and the barrel of the cylinder with the upper and lower caps only re- 
tained. This is installed in the air line, as shown in Figure 7, between 
the valves U and W . A brass cup of heavy gauge metal (Fig. 6 A) was 
made to slip inside this cylinder barrel. It has a ^4 -inch flange at the 
top, resting on the upper end of the cylinder barrel under the leather 



LABORATORY METHODS OF' TESTING SANDS 47 

gasket, insuring an air-tight joint. The bottom of the cup is heavy stiff 
brass, perforated. All of the samples were mounted in this cup, the 
method being adapted to the nature of the sample. 

When damp or green sand is tested the sample is rammed into a 
special cylinder (Fig. 6 B and C) of the same dimensions used in the 
porosity tests, viz., 2 13/32 inches in diameter by 3 inches long, open at the 
ends and having a flange ^4 of an inch wide at the upper end. The flange 
rests on top of the cup beneath the leather gasket, thus insuring an air- 
tight joint, and compelling all of the air to pass through the sand core. 
The lower end of the sand core in this cylinder is supported by building 
up the bottom of the brass cup with discs of brass wire screen cloth of 
20-mesh. 

When loose sand was tested, the same cylinder (Fig. 6 C) was used. 
To retain the sand a false bottom of screen cloth was slipped on over 
the cylinder. The loose sand was. compacted in the cylinder by rapping 
the side of the cylinder to jar the contents. The cylinders of loose sand 
were mounted in the brass cup in the same manner as the damp sand. 

Where dry cores of sand (Fig. 6 F) were to be tested, it was neces- 
sary to provide some envelope that would fit the side walls of the core 
tightly enough to prevent any air by-passing along the sides of the core 
between the core and the envelope. Two types of envelopes were used 
to accomplish this purpose, and the cores were mounted in the brass 
cup as follows : 

1. The cores by one method are coated with melted paraffin, care be- 
ing exercised to have the paraffin no hotter than needed to apply it. If 
paraffin is too hot and fluid, it permeates the core and diminishes its per- 
meability. After coating with paraffin, the core is set in the center of 
the brass cup, and the space between the core and the sides of the cup, 
about one half inch, is filled with puddled clay in a semifluid condition. 
The bottom of this space is filled with cotton waste and a little sand, to 
prevent the fluid clay passing through. This type of mounting was found 
to be absolutely impermeable. 

2. The second type of envelope used for the dry cores consists of a 
sleeve of very thin sheet rubber, such as dentists use, which they call 
"rubber dam." These sleeves are made with a diameter less than the 
diameter of the core, vulcanizing the seam. The sleeves are stretched 
and slipped over the cores. The cores with rubber envelopes are mounted 
in the brass cup in the manner above described, using puddled clay in 
the intervening space. Paraffin was tried instead of puddled clay, But 
with unsatisfactory results, since the paraffin on cooling would shrink 
and draw away from the walls of the cup. 



48 THE FOUNDRY SANDS OF MINNESOTA 

It is well known that to devise an apparatus that will deliver a definite 
volume of air at a constant pressure is a most difficult problem. The 
most serious fault with the apparatus used by King was that the volume 
of air employed was small and the pressure low, and because of this the 
barometric pressure with its diurnal variation, the humidity of the air, 
and the temperature of the air became important variable factors that 
had to be taken into account and determined with each run made. By 
greatly increasing both the volume of air passed and the pressure, these 
variables became practically negligible. 

In making the tests of the permeability of the molding sands a record 
was kept systematically of the barometric pressure, temperature, and 
humidity of the air to determine to what extent these variables were 
factors in the problem ; but it was found that the corrections necessitated 
on this account were so slight that they could be ignored without appre- 
ciably affecting the results. 

RELATION OF POROSITY, TEXTURE, AND STRUCTURE TO 

PERMEABILITY 

In ordinary sand and loam such as are used in foundries the inter- 
stitial spaces between the sand grains, called the voids, communicate with 
one another, forming continuous tube-like passages which the air in pass- 
ing through the sand follows. They are very small, hair-like, in size. 
The movement of air through sand has been demonstrated by laboratory 
experiment to be comparable, within certain limits, to the movement of 
air through capillary tubes, indicating that the interstitial passages 
through sand are capillary in size. 

The size of the interstitial tubes in sand is a function of the size and 
arrangement of the individual grains. Slichter has shown that in a. mass 
of perfect spheres arranged in the most compact form, shown in Figure 
8, the area of the cross section of an interstitial tube through the mass is 
.1613 r 2 , in which r is the radius of the individual sphere. 8 

The rate that air will pass through capillary tubes of different diam- 
eters and the same length is in proportion to the square of the area of 
the cross section of the respective tubes. With two capillary tubes the 
areas of the cross sections of which are 2 and 4 square microns re- 
spectively the rate of the passage of air through them would be in the 
ratio of 4 to 16. From which it will be seen that the size of the inter- 
stitial tubes in sand is of vital importance as effecting the permeability. 

Texture in sand is that property which is related to, and dependent 
on, the size of the individual grains. For example, a sand in which the 

8 Slichter, C. S. Theoretical investigations of the movement of ground water: U.S. Geol. 
Survey Nineteenth Ann. Rept., part 2, p. 316, 1899. 



TEXTURE, STRUCTURE, AND PERMEABILITY 



49 



grains were all of the 20-mesh size, or in which the 20-mesh sand pre- 
dominated, would be said to have a coarse texture; whereas a sand in 
which all of the grains were of the 80- or 100-mesh size, or in which 
these sizes predominated, would be said to have a fine texture. Struc- 
ture in sand has reference to the arrangement of the grains with respect 
to each other without regard to their size. 




FIGURE 8. DIAGRAM SHOWING THE CLOSEST SYSTEM OF PACKING THAT IS POSSIBLE 
FOR SPHERES OF UNIFORM SIZE. THE RODS INDICATE LINES OF FLOW 

AFTER SLICHTER 



The permeability of sand is dependent on each of these properties, 
the relative importance of which varies with the individual case. Ref- 
erence to Figures 9, 10, and 11 may serve to make clear why this is true. 

Figures 9A and 9B illustrate the effect of texture on permeability. 
The structure is the same in both 9A and 9B ; that is, the spheres are 
systematically placed in the most open arrangement with each sphere 



50 



THE FOUNDRY SANDS OF MINNESOTA 



touching six other contiguous spheres. The spheres in Figure 9B, how- 
ever, are just one half the diameter of the spheres in Figure 9A, and 
accordingly there are four times as many interstitial tubes in the mass 
9B, as there are in the mass 9A. The total interstitial space, or porosity, 




FIGURE 9A AND 9B. ILLUSTRATING 
STRUCTURE IN MASS OF SPHERES 
IN MOST OPEN ARRANGEMENT IN 
WHICH VOIDS EQUAL 47.6 PER 
CENT OF THE VOLUME OF THE 
MASS 

After F. H. King 




F>9- 10 A 




Fi'S /OB 



FIGURE IOA AND ICB. ILLUSTRATING 
STRUCTURE IN MASS OF SPHERES OF 
SAME SIZE AS FIGURE Q IN MOST 
COMPACT ARRANGEMENT, IN WHICH 
THE VOIDS EQUAL 25.9 PER CENT OF 
THE VOLUME OF THE MASS 

After F. H. King 



is the same in both 9A and 9B, viz., 47.6 per cent. Therefore the inter- 
stitial tubes in the mass 9A must be four times as large as the interstitial 
tubes in the mass 9B. The areas of the cross section of individual tubes 
in the masses 9A and 9B would be in the ratio of 4 to 1 ; but according 
to the law stated in the preceding page the rates at which air would pass 



TEXTURE, STRUCTURE, AND PERMEABILITY 



5i 



through these individual tubes in masses 9A and 9B would be- in the 
ratio of 16 to I. Since there are four times as many tubes in 9B as 
in 9A the rate at which air would pass through the two masses 9A and 
9B would be in the ratio of 4 to 1 ; that is, the mass 9A would be four 
times as permeable as the mass 9B, notwithstanding the porosity is the 
same and the structure is the same in the two masses. The only differ- 
ence in the two masses is that 9B is of finer texture than the mass 9A. 







^^^^^^S 



WK^SSKKKKKKB^KKKKB^^^ 



FIGURE II. ILLUSTRATION OF RHOMBOHEDRON CUT FROM MASS OF SPHERES IN MOST 

COMPACT ARRANGEMENT, SHOWING ARRANGEMENT, RELATIVE SIZE, AND 

COMMUNICATION OF INTERSTITIAL SPACES 

AFTER SLICHTER 

The size of grain is the same in Figure 9 A and in Figure 10A, but 
the arrangement of grains, or structures of the masses is different. The 
spheres shown in Figure 9A are placed in the most open arrangement 
possible, with each sphere touching six other spheres, 9 whereas the spheres 
shown in Figure 10A are in the most compact arrangement possible, with 
each sphere touching twelve other contiguous spheres. The porosity of 
the mass 9A is 47.6 per cent; that of 10A is 25.9 per cent. There are 
sixteen interstitial spaces in 9A and thirty-two in 10 A, with an equal 
number of sand grains. The ratio of numbers of pores is therefore 
about 1 to 2. 

If one pore in 9A had an area of 47.6 units, a single pore in 10A 
would have an area of about one half of 25.9 units, or 12.95 units, 
roughly one fourth the area of the pores in 9A. The permeability vary- 
ing as the square of the area of a pore (here a tube) would be 16 times 

9 Four spheres only are shown in the figure which is a cross section but two other spheres 
are not shown; one is at the front, another at the back of section. In figure 10A only six spheres 
are shown of twelve that touch each sphere. 



52 



THE FOUNDRY SANDS OF MINNESOTA 



as great in each tube of o,A as for each tube of 10A. But there are about 
twice as many tubes in 10A, so that the permeability of a mass like 9A 
will be roughly 8 times as great as that of 10A. This great difference 
in permeability is independent of the size of sand grain used, so long 
as they are uniform spheres. The difference is wholly a matter of ar- 
rangement of grain or structure in the two masses. 

Another illustration of the effect of structure on permeability is shown 
in Figure 12, where the intergranular spaces. _ rather than the interstitial 
spaces determine the permeability of the niass. 




FIGURE 12. ILLUSTRATION OF GRANULATION IN 
WHICH SPHERICAL GRAINS ARE BOUND TOGETHER 
IN COMPACT GRAINS OR GRANULES ARRANGED IN 
OPEN STRUCTURE WITH LARGE INTERGRANULAR 
SPACES 

AFTER LYON, FIPPEN, AND BUCKMAN 

Each of four granules A, B, C , and D, is made up of spherical grains 
placed in the most compact arrangement, and the granules in turn are 
arranged in open structure so that each granule touches six other gran- 
ules, leaving relatively large intergranular spaces. In a structure of this 
kind the spaces between granules or groups of spheres alone would 
amount to more than 40 per cent of the mass ; in addition to which other 
spaces between spheres amounting to about 15 per cent of the entire mass, 
would give a total of about 55 per cent of voids. The smaller tubes 
between grains in the granules would be so small, however, compared 



EFFECTIVE SIZE OF GRAIN 53 

with the intergranular tubes, that they would have little effect on the 
permeability. Owing to the large intergranular tubes the permeability 
of the mass would be very great. 

The case illustrated in Figure 12 is hypothetical. In natural sands the 
grains are never perfect spheres, nor are they ever symmetrically ar- 
ranged. On the other hand in molding sand that has been tempered with 
the proper amount of moisture, granulation does occur and the permea- 
bility of such tempered sand is usually greatly increased as a result of 
the granulation. In soils these clusters or compound grains have been 
termed the ''effective grain" because they behave as unit masses and 
give to the soil a structure and a permeability the same as though the 
soil were made up of particles of this size. In soils or in molding" sand, 
however, the compound grains or granules may be made up of hundreds 
or even thousands of individual particles instead of a few particles as 
illustrated in Figure 12. 

As to the effect of granulation on permeability in molding sands a 
single illustration will suffice. In Table XVI, it will be observed that 
with sand 240, in the loose dry state in which the sample had been pestled 
to break up the compound grains into individual particles, it required 
2,027 seconds to pass a given volume of air through a standard cylinder 
of the dry sand, whereas the same sand moistened, tempered to facilitate 
granulation, and then rammed into the same cylinder allowed the same 
volume of air to pass in about one tenth of the time, viz., 244 seconds. 
The greater permeability of the same sand in the damp molded condi- 
tion was plainly the result of granulation. 

EFFECTIVE SIZE OF GRAIN 

In natural sands of unassorted sizes the interstitial spaces between 
the sand grains vary greatly in size and are more or less haphazard in 
arrangement, but it has been demonstrated mathematically, and confirmed 
by laboratory experiment, that with clean loose sand grains of mixed 
size there is a mean, or average grain, called the "effective size" ; and that 
the permeability of such sand is the same as though the mass were made 
up entirely of grains of this effective size. It has been shown also that 
within certain limits the movement of air through sands made up of 
grains of mixed sizes is amenable to the laws governing the movement 
of air through masses made up of grains spherical in form and sym- 
metrical in arrangement. 

It has been shown by Slichter that in masses made up of spherical 
grains in most compact arrangement the area of the cross section of in- 
terstitial tubes is .1613 r 2 , in which r is the radius of the individual sphere. 
But because of the irregularity in outline of the cross section of such a 



54 



THE FOUNDRY SANDS OF MINNESOTA 



tube, not all of this area is to be counted as effective in the passage of 
air. Making allowance for this factor, Slichter computes the effective 
area of cross section of interstitial tubes in granular masses such as above 
described to be .1475 r 2 , in which r is the radius of the individual sphere 
in the mass. 



■lb 

J3 
Jl 

I 



£.09 

i 

t 

8: 

N 

<o 

"5 
$..03 

u 

.01 


































































































































































.00 













(60 



/20 100 SO 

Mesh of Screen 



60 



40 



20 



4- 



FIGURE 13. 



DIAGRAM SHOWING RELATIONS OF MESHES OF SCREENS AND 
ACTUAL SIZES OF OPENINGS 



A determination of the size of the granules, or compound grains, was 
not attempted because numerous factors were involved, the relative values 
of which have not been determined. The laboratory work was devoted 
largely to special problems submitted by foundries, and a determination 
of the actual and comparative permeability as measured by the time re- 
quired to pass a standard volume of air through particular samples of 
sand was all that was necessary. 

METHOD OF COMPUTING PERCENTAGE PERMEABILITY 

For comparing the permeability of sands, it is convenient to have 
some factor other than the time of passage of the air to express the rela- 
tive permeability of different sands. Accordingly standard Ottawa sand 
was taken as a standard for comparison, and assumed to be 100 per cent 
permeable. The standard sand used in these experiments, sample 215, 
showed by screen analysis less than 1/10 of 1 per cent coarser than 
20 mesh and less than 1 per cent finer than 30 mesh. 



COMPUTING PERCENTAGE PERMEABILITY 55 

In the permeability apparatus sand 215 showed as a result of seven 
tests that it required 24 to 26 seconds to pass the standard volume of 
air, 74 liters, through this sand, giving an average of 25 seconds. As- 
suming this sand to be 100 per cent permeable and 25 seconds to be the 
time required to pass the standard volume of air through it, then the 
percentage permeability of the other sands was calculated by the formula 
25 = Z, in which Y is the time required to pass the standard volume 
Y~ 

of air, 74 liters, through the sand whose permeability is sought, and Z 
is the percentage permeability. For example, with sample 9, it required 
94 seconds to pass the standard volume of air through the sand,, accord- 
ingly 25 = 26.6 per cent, and 26.6 is taken as the per cent permeability 

94 

of sand 9. Sample 240A required 2,027 seconds to pass the standard 

volume of air, and 25 = 1.2 per cent is the permeability of this sand. 

2027 

In making the permeability tests the same practice was followed as 
in making tests for bond, i.e., such samples as were molded were made 
in triplicate, and the average taken of three tests. A core box was used 
in most cases, permitting the making of three cores at a time. Where 
possible each core was run successively for green, for dry, and for baked 
test; that is, the core was put in the permeability apparatus directly from 
the mold and tested for permeability while damp or green. It was then 
allowed to stand in the laboratory 24 to 48 hours to dry, after which it 
was run again as an air-dried core. The same core was then placed in 
the bake oven at 400 C. for 4 to 8 hours, after which it was allowed 
to cool and was tested again for permeability as a baked core. 

It frequently happened that the cores of sand molded directly in the 
solid cylinders became loose in the cylinders after drying, as a result of 
shrinkage from loss of moisture. In such cases the cores were removed 
from the cylinders and provided with a special mounting, using a seal of 
melted paraffin or puddly clay to insure against leakage of air around 
the sides of the core, and compelling all of the air to pass through the 
core itself. 

The apparatus was so arranged that the pressure could be brought 
to the standard amount, namely, 50 inches water, and held there for any 
period desired. By observing the manometer and the water gauge, any 
leaks in the mounting: were at once detected and corrected before making 
the run. 

•ill. 1.. vi ii; .1 1 



56 THE FOUNDRY SANDS OF MINNESOTA 

SHAPE AND ARRANGEMENT OF INTERSTITIAL TUBES IN SAND 

The shape and arrangement of the interstitial spaces or voids in a 
mass of sand, and the manner in which these voids communicate to form 
interstitial passages or paths through the mass is illustrated in Figures 
8 and 1 1. The grains of the mass in these illustrations are perfect spheres 
most compactly arranged, in which the pore space is about 26 per cent. 
It will be observed that the interstitial tubes through the mass, indicated 
by the rods, in Figure 8 are inclined at an angle of 30 from the vertical. 
They form passages following rectilinear paths through the mass, and 
as Slichter has shown the length of these passages or interstitial tubes 
is 1.065 times as great as a direct line through the mass normal to the 
base or top of a sand column. Accordingly in the cores of sand used 
in the permeability experiment, which were 3 inches long, the actual 
length of the interstitial tubes through these cores would be about 3.195 
inches. 

In the flow of air through capillary tubes the length of the tube as 
well as its diameter is a function of the rate of flow. The formula for 
computing the flow of air through capillary tubes was worked out ex- 
perimentally by Poiseuille in 1842, and is known as Poiseuille's Law. 
The proper ratio of the length of a capillary tube to its diameter in order 
that Poiseuille's Law shall hold is about 6,000 to 1. This should be taken 
into consideration if the results given in Table XVI are to be checked 
against Poiseuille's Law. 

Cores used in permeability tests of sands, loams, and clays. — The 
purpose of the permeability test was to obtain actual figures on the 
comparative permeability of different kinds of sand under conditions 
comparable to those in the foundries rather than as a study in per- 
meability per se. The actual dimensions of the cores as finally chosen 
were determined by several considerations as follows : 

1. The depth of the core, or length of the sand column, should be 
comparable to the thickness of the walls of molds in ordinary foundry 
work. Three inches was thought to be a fair average thickness for 
such molds, and accordingly a core 3 inches deep was chosen. 

2. The structure in a sand mass adjacent to, or in contact with, the 
walls of the container is necessarily somewhat different from what it 
is in the center of the mass. This is especially true in damp sands that 
are rammed into the container, because of the drag of the sand on the 
side walls. Accordingly the larger the diameter of the container the less 
serious this factor becomes. 

3. The air passages through a sand tend to follow rectangular paths and 
some of these paths meet the side walls and have their courses deflected. 
The larger the diameter of the container, the less the percentage of the air 



MOISTURE OF MOLDING SANDS 57 

: passage that will be interrupted in this manner, and the more nearly 
1 the test cylinder approaches normal conditions. 

4. A sand column the length and diameter of which are the same 
will probably most nearly approach foundry conditions, and accordingly 

; a core was sought as near 3 inches in diameter as practicable. 

5. It is desirable to use brass cylinders of heavy gauge so that the 
i walls are not deformed, or the cross section of the sand column distorted 
I by ramming sand into the cylinders, or by air pressure. Furthermore 

the plan of the work contemplated a large number of these cylinders for 
use in the foundries, and it was essential that they be as near the same 
dimensions as possible. It was found that common rolled brass tubing 
i met the requirements, but such tubing 3 inches in diameter was not 
1 readily available and tubing 2 13/32 inches in diameter was chosen as 
I the nearest approach to the dimensions desired. The containers were 
l made by cutting sections 3 inches long from such tubing. 



FUSION TESTS 

The fusion tests of sands consisted in burning two or more briquets 
1 of the various sands in an electric furnace, the temperature of which 
1 was controlled, and noting the effect of temperatures ranging from 400 
1 to 1,000 degrees C. The furnace was provided with a window or peep- 
i hole, so that the behavior of the samples could be observed as the tem- 
perature was increased. Seger cones were used in the furnace along 
with the samples, following the usual practice in ceramic work, as a 
check on the furnace temperature. The laboratory work in fusion tests 
1 was not sufficiently complete to be made the basis of generalizations, but 
1 was suggestive in that some of the sands that had ample bond, and gave 
satisfactory results in the foundry, disintegrated into loose sand in the 
furnace at temperatures of 600 to 1,000 degrees C, whereas other sands 
i burned into fairly hard brick at the same temperatures. In general the 
1 temperatures reached were not high enough to cause even incipient 
fusion, although in a few cases the stage of vitrification was reached. 

MOISTURE CONTENT OF MOLDING SANDS 

Certain tests for moisture content of molding sand were undertaken 
} with the following objects in view: (a) to determine the moisture con- 
! tent in sands and loams in actual use in the foundries; (b) to determine 
1 whether or not better results in tensile strength or permeability might 
i be obtained by increasing or decreasing the amount of moisture from 
1 that used in the foundries; (c) to determine the variation in the moisture 
in the molding sands in the same foundry from day to day. 



38 THE FOUNDRY SANDS OF MINNESOTA 

The results obtained served to show that there was a considerable 
range in the moistur content in the same foundry from day to day in 
the same kind of molding, the variation amounting to 30 per cent in some 
instances. There was also a variation in different foundries using the 
same grade of sand, and doing the same class of founding. 

There was a variation in the moisture content in the same foundry 
as between different molders using the same sand. The results are re- 
corded in the tables on tensile strength and permeability and are discussed 
in that connection. 

The method of taking samples for moisture determinations was to 
use two quart glass fruit jars with the ordinary rubber gasket to insure 
against loss of moisture. The samples were taken from the moldefs' 
tables or from the sand heaps in actual use. r - 

The method of determining the moisture content consisted in drying 
100-gram samples of the sand taken in the foundry in the sand bath at 
100 degrees C. for 24 hours or longer until they ceased to lose weight, 
and counting the loss of weight as moisture. 

MECHANICAL AND MINERAL ANALYSES OF 
CORE SANDS 

Results of the mechanical analysis of core sands are recorded in 
Table IX. Forty samples were analysed, following both the dry method, 
A, and the wet method, B, above described. The analyses by the dry 
method, A, are not recorded except in cases where the wet method, B, 
was not completed. The 19 analyses recorded in Table IX were selected 
as being typical of the total number made. 

Eight of the analyses given in Table IX are of samples obtained 
in the foundries of sands in actual use in core work. The remaining 
samples, the analyses of which appear in Table X, were collected in the 
field from various parts of the state as representing material similar 
in character to that in use for core work. 

One sample, 252A, used for core work in exceptionally heavy gray 
iron work, was obtained by the foundry from local glacial drift, having 
been washed at the pit to remove the greater part of the clay, and screened 
to remove the gravel coarser than J4 mcn m diameter. The other samples 
of core sand in use in the foundries were obtained from the St. Peter 
formation in St. Paul and Minneapolis. 

The analyses, Table IX, show that sand suitable for core work may 
be obtained from the St. Peter formation carrying less than one half 
of 1 per cent of clay, but most of the sands in actual use carry 1 to 2 
per cent of clay. 



ANALYSES OF CORE SANDS 59 

The texture of the core sands in actual use in the foundries is indi- 
cated in a general way by the "mesh" recorded in Table IX. The range 
for samples of sand used, except 252, is from 52 to 72. Sample 5, with 
a comparative mesh of 52, is the common type. The method of computing 
the comparative mesh, and the purpose in using the term to express texture 
is explained above, under Laboratory Methods. Sample 90, showing a 
mesh of 38, was the coarsest sand found in the St. Peter formation. 

For ordinary core work the St. Peter sand is very satisfactory, and 
there is an unlimited supply available. The strictly clean sand is not 
necessarily the best. Samples 28 and 31 were used in the same foundry, 
and both came from the St. Peter formation. Number 28 was a clean, 
white, quartz sand, whereas 31 was stained yellow by oxide of iron. 
The yellow sand was the more satisfactory, as it developed greater tensile 
strength and gave smoother surfaces to the castings. 

Where coarser sand was desired for heavy core work, this was ob- 
tained by mixing sand from the Jordan formation, described under steel 
sand, with the St. Peter sand. Coarser sand is obtainable from the St. 
Peter formation, but it is usually present as comparatively thin horizons, 
of 2 to 6 feet, interbedded with sand of the finer type. It is not feasible 
in most instances to develop it because of the overburden of fine material, 
and because of the excessive expense of mining it by tunneling. It is 
cheaper ordinarily to purchase the coarser sand from the Jordan forma- 
tion and raise the average mesh by doctoring the St. Peter sand. 

Owing to the rather wide variation in the requirements of core sand 
for the different classes of founding, no single grade of sand meets all 
of the needs. For core work in brass, and light work in grey iron, a 
sand of about 57 average mesh is preferred, and for heavy work in grey 
iron and steel an average mesh of about 35. In general, while a clay 
content is not essential, 1 per cent of clay or even more, is no detriment. 
The entire absence of silts is preferred, but the presence of 1 to 5 per 
cent of silt does no apparent harm. 

Samples 90, 145, and 278 were collected from the St. Peter forma- 
tion at Cannon Falls, Rochester, and Northfield respectively and illus- 
trate the uniformity of the sand in this formation throughout the state. 

Sample 72 was collected at Daytons Bluff, St. Paul, from the St. 
Peter formation, and illustrates the type of very fine sand that occurs 
in the St. Peter formation at certain horizons. 

Samples 152 and 279 were collected from pockets and lenses of sand 
that occur more or less commonly in the dolomitic limestone between 
the Jordan and the St. Peter formations. 

Sample 104 was collected from the Cretaceous formation in Goodhue 
County. It is of no commercial importance. 



6o 



THE FOUNDRY SANDS OF MINNESOTA 



The St. Peter and Cretaceous sands are largely quartz. Number 252 
is a glacial sand and has about 22 per cent of minerals other than quartz. 
As it is being used successfully, it would indicate that mineral composi- 
tion is of little importance. From the standpoint of refractoriness, those 
highest in quartz should be best, but actual practice shows that others 
can be used with a fair degree of success. In the northern part of Min- 
nesota, glacial sands are used considerably for core work. One from 
Biwabik is used in the Virginia Foundry, and one from Brainerd is used 
by the Brainerd Foundry. Lake Superior sand is used for core work 
at Duluth to some extent. 



TABLE IX. MECHANICAL ANALYSES OF CORE SANDS 



Mesh of Screen 


iB 


SB 


21B 


28B 


3iB 


64B 


65A 


67A 






1.40 

1.48 

56.66 

22.86 

11.62 

5.26 

.07 

Trace 

Trace 

• 05 


1.04 

41.38 

17.32 

13.96 

23.00 

1.2.6 

Trace 

Trace 

1.44 


1.22 

54-68 

22.10 

12.32 

8.42 

.24 

.00 

.00 

.28 


1.02 

40.58 

18.94 

14.24 

21.66 

1.36 

Trace 

1. 00 

.46 


48.74 

39.56 

3-68 

1.92 

3.76 

.96 

Trace 

.16 

.52 


9.87 

58.76 

7.92 

7.61 

12.92 

.26 


Trace 


20 — 40 


.86 

16.24 

34-86 

21.92 

22.56 

1.04 

.10 

.20 

2.16 


40.27 

40.45 
4.60 


60 — 80 


80 — 100 


3-02 

7-37 
2.42 


100 — 200 


200-silt 


15-minute silt 




30-minute silt 




Clay 








Total 


99-94 

72 


99.40 

52 


99.40 
59 


99.26 

55 


99.26 
59 


99-30 
39 


97-34 
5i 


98.13 
41 





1. Core sand, University of Minnesota, foundry, from St. Peter formation. 
5. Core sand, Eagle Foundry, Minneapolis, used in grey iron work, from St. Peter 
formation. 

21. Core sand, American Hoist & Derrick Foundry, St. Paul, used in grey iron work, 
from St. Peter formation. 

28. Core sand, Northern Malleable Iron Works, St. Paul, from St. Peter formation. 
31. Core sand, Northern Malleable Iron Works, St. Paul, from St. Peter formation. 

64. St. Peter formation, Forty-second and Randolph Streets N.E., Minneapolis, four feet 
below top of formation. 

65. St. Peter formation, same locality as No. 64, twenty feet below top of formation. 
67. St. Peter formation, Washington Avenue N. and Thirt'y-ninth Street, Minneapolis. 



ANALYSES OF CORE SANDS 



61 



TABLE IX— Continued 



Mesh of Screen 



4 — 10 

10 — 20 

20 40 

40 — 60 

60— 80 

80 100 

100 200 

200-silt 

15-minute silt . . . 

30 -minute silt' . . , 

Clay 

Total 

Comparative mesh 



72B 



.16 
.80 
.90 
3.00 
32.98 
59.52 
Trace 
Trace 
1.76 
99.12 
230 



90B 104B 145B 152A 152B 



57.00 

29.20 

1.84 

I.IO 

.94 

.52 

.00 

2.48 

2.26 

95-34 

38 



3.12 
33-90 

27.36 

12.50 

12.02 

3.20 

.60 

.36 

5.62 

98.68 

60 



48.92 

28.76 

6.28 

3-04 

5-40 

5.14 

Trace 

1.84 

3.40 

102.78 

41 



20.36 
21.68 
33-20 
10.68 
6.65 
6.54 



99.11 

52 



I52.B 


252A 


Sand 


2.28 


not 


14.76 


screened 


55.64 


total 


13-54 




13.18 




87.00 


2.46 




1. 16 




.70 


1. 00 




5.00 





5-74 




98.74 


103.72 




27 



262 A 



2.2.6 
45.i6 
11.96 
10.02 
28.84 

1.06 



99-30 

57 



72. 
90. 
104. 
145. 
152. 
2.52. 
washed. 

262. 
formation. 



St. Peter formation, Daytons Bluff, St. Paul. 

St. Peter formation, Cannon Falls, Minn. 

Cretaceous formation, four miles southwest of Goodhue, Goodhue County, Minn. 

St. Peter formation, Klines' quarry, Rochester, Minn. 

Oneota formation, sand bed in limestone, Orinoco, Olmstead County, Minn. 

Core sand, Eagle Foundry, Minneapolis, used in heavy core work from the glacial drift, 

Core sand, Soo Line Railroad Shops, used in brass founding, from St. Peter 



Mesh of Screen 


27 8B 

3-00 
22.00 
26.52 
18.00 
21.28 

5.04 
1.06 
1.22 
1.40 


279B 


282B 


299A 


299B 


4 — 10 


14.82 

57.62 

16.30 

5-oo 

4-54 

2.04 

.16 

.64 

1.82 


.42 

29.20 

16.54 

13.84 

31.58 

2.30 

.20 

.20 

2.26 


.90 
48.94 

19.44 
14.04 
15-50 

I.IO 




10 — 20 


Sand 




not 


40 — 60 


screened 


60 — 80 


total 


80 — 100 






98.54 


15-minute silt 


.00 


30-minute silt 


.18 


Clay 


1.58 


Total 


99.52 
68 


102.94 
49 


96.54 
68 


99.92 

57 


100.30 



278. St. Peter formation, Northfield, Minn. 

279. Shakopee formation, Cannon Falls, Minn., sand in limestone. 
282. St. Peter formation, Twin City Brick Co., St. Paul. 

299. Core sand, Flour City Ornamental Iron Works, foundry, Minneapolis, from St. Peter 

formation, used in brass and bronze ornamental founding. 



62 



THE FOUNDRY SANDS OF MINNESOTA 



MINERAL ANALYSES OF CORE SAND 



Number 



Part examined (over 200 mesh) 

Quartz 

Tourmaline 

Limonite 

Kaolin 

Dolomite 

Feldspar . 

Igneous rock 

Hornblende . . 

Chlorite 

Augite 

Magnetite 

Biotit'e 



64 


104 


278 


97-66% 
99.2 

.02 

.01 


9-2.4% 
98.97 
.01 
.18 


90.8% 

99-3 

.07 




.26 













































252 



.100% 

76.16 

.01 

•03 



9.41 
7. II 

3-28 

1. 41 

1.2-7 

.II 

.01 
.01 



MECHANICAL AND MINERAL ANALYSES OF STEEL SANDS 

The results of analyses of the coarser quartz sands, known as steel, 
fire, and silica sand, are recorded in Table X. The chief difference 
between these steel sands and the core sands is in the size of the grains, 
that is the steel sands are coarser than the core sands as generally under- 
stood ; but there are exceptions to this rule, for in very heavy work core 
sands are used quite as coarse as the so-called steel sands. Fire or silica 
sands are quartz sands of medium to coarse grain that are very refrac- 
tory; that is, their ability to stand heat is equally as important as their 
texture. They may contain 2 per cent or more of clay, as does sand 9, 
but if so this clay also must be refractory. 

Most of the silica sands, or steel sands used in the Minnesota foun- 
dries are obtained from the Jordan formation, and would probably show 
on chemical analysis 92 to 95 per cent silica. The St. Peter formation 
carries some coarse sand also, as shown above in Table IX, which is 
essentially the same as the Jordan sand ; in fact the sized sands from 
the two formations can not be distinguished from one another. The 
quartz grains, particularly the larger ones, are exceptionally well rounded. 

Samples 166, 167, 168, 170, 173, and 175, were collected from the 
Jordan formation in the valley of the Minnesota River between Mankato 
and Merriam Junction, while samples 8 and 19 were collected from the 
foundries, and came originally from this same region. The similarity in 
the analyses is apparent. There is a noticeable difference in the texture 
of the Jordan sands, the comparative mesh ranging from 28 to 42. 

The wet analysis to determine the clay content was not made on all 
of the steel sands, but analyses 8 and 19, which are wet analyses, are 
representative. 

Sample 136 was collected from the Jordan formation 10 miles north 
of the Iowa line, at Preston, Fillmore County, about 100 miles southeast 



ANALYSES OF STEEL SANDS 63 

of the area in the Minnesota River Valley. It is of no commercial value, 
except locally, but is of interest as showing the nature of the Jordan 
sand in this part of the state. This sample was not particularly coarse, 
but was practically free of silt and clay. 

Sample 151 was collected from the Jordan formation 35 miles north 
of the location just mentioned, at Orinoco, Olmstead County. It differs 
from 136 in being coarser and in carrying 2 per cent of clay and 3 per 
cent of silts. Samples 136 and 151 resemble the St. Peter sands. 

Sample 160 was collected from the Jordan fomation in the valley of 
Zumbro River, 2 miles below Mazeppa, Wabasha County, 12 miles north 
of Orinoco. In coarseness and in the content of clay and silt it resembles 
the Jordan sands, 75 miles to the westward in Scott and Le Sueur coun- 
ties, in the Minnesota River Valley. 

Sample 9, from Massilon, Ohio, differed from the local sands chiefly 
in having a slight matrix of semi-refractory clay that formed a very 
uniform coating over the sand grains, which was sufficient to mold the 
sand. 

Samples 185 and 186 were collected at Sandstone, Pine County, 80 
miles north of St. Paul, from the Kettle River formation. The Kettle 
River sandstone is extensively quarried at this point for building stone 
and the rubble is crushed for concrete aggregate. The samples represent 
the sand screenings from the crushers, which has been used in foundries 
as steel sand with satisfactory results. This sand is rather finer than 
the Jordan sand, and carries more clay and silt. 

Sample 188 was collected from the Sioux quartzite, Nicollet County, 
in the quarries near Courtland below New Ulm. The quartzite here 
along the quarry joints has disintegrated under the influence of organic 
acids and has reverted to sand. It is of no commercial importance, but 
is of interest as showing the character of the original sand from which 
the Sioux quartzite was formed. 

Sample 200, Table XIV, while not listed as steel sand, might be used 
to advantage for that purpose. It consists of screenings from the crush- 
ers at Jasper, Rock County, Minnesota, where the Sioux quartzite is 
crushed for concrete aggregate. It is practically pure quartz, very re- 
fractory, nearly free from clay and silt, and extremely angular. 

Steel sand is imported from Ottawa, Illinois, for use in the foundries 
of Minneapolis and St. Paul, and while no analysis of this sand appears 
in Table X, this Illinois sand was tested and found to be identical in 
physical composition with several samples of the Jordan sand of 
Minnesota. 

In this connection attention should be called to a confusion of terms. 
Some founders understand the term Ottawa sand to mean sand from 



64 



THE FOUNDRY SANDS OF MINNESOTA 



Ottawa, Illinois, while others understand it to mean sand from Ottawa^ 
Minnesota. That the two towns happen to have the same name is a 
coincidence only. The sands are essentially the same. The Ottawa, 
Illinois, sand, however, happens to come from the St. Peter formation, 
whereas the Ottawa, Minnesota, sand comes from the Jordan formation 
some 200 feet below the St. Peter formation. 

The requirements for steel sand are somewhat varied and for this 
reason no one grade meets all needs. For some purposes a content of 
refractory clay sufficient to mold the sand is desired and if this can be 
obtained in the natural state it is preferred. The Kettle River sand comes 
near meeting these requirements ; the sand from the Jordan formation 
rarely does. 

Since it is desired that steel sands show a silica content of 97 per cent 
or over, 10 the analyses indicate that, the Jordan, Kettle River, and Sioux 
are satisfactory from a mineral standpoint for such use. 

The comparative mesh of steel sand is about 42. The absence of silts 
in steel sand is preferred, but 1 to 3 per cent of silt is not a serious 
detriment. 

TABLE X. STEEL AND FIRE SANDS 



Mesh of Screen 



4 — 10 

10 — 20 

20 — 40 

40 — 60 

60 — 80 

80 — 100 

100 — 200 

200-silt 

15-minute silt 

30-minute silt 
Clay 



Total 

Comparative mesh 



8B 


9B 


19B 


136A 


151A 


151B 


I59B 


3-44 


•74 


4.02 




.36 


Sand 


3.08 


54.8o 


22.80 


44-47 


17.40 


0.00 


not 


68.40 


29-54 


26.06 


27.08 


28.88 


20.02 


screened 


23-52 


4.68 


25.08 


5.56 
3-96 


31.96 
15-52 


36.40 
13.72 


total 


2.52 
.80 


2.30 






2.61 


7-44 


9-25 


5.42 


16.88 


95-oo 


.42 


• 30 


3-50 


3-8i 


.28 


1 a. 44 




.02 


Trace 


.12 


.63 






1. 00 


.00 


Trace 


1.40 


.10 






2.00 


.00 


.25 


2.00 


.98 






2.00 


.46 


97.92 


99.14 


99.86 


99.46 


99.82 


100.00 


99.22 


35 


50 


40 


5i 


7i 




32 



I60B 



1. 18 
48.58 
35.68 

9.04 
2.94 

.98 

.00 

Trace 

.24 
.92 

99-56 
38 



8. Steel sand, Gas Traction Foundry, Minneapolis, from Jordan formation, Ottawa, Minn. 

9. Steel sand, Gas Traction Foundry, Minneapolis, from Massilon, Ohio. 

19. Silica sand, American Hoist & Derrick Foundry, St. Paul, from Ottawa, Minn.', 
Jordan formation. 

136. Quartz sand, Jordan formation, at Preston, Fillmore County, Minn., 10 miles north 
of Iowa line. 

151. Quartz sand, Jordan formation, at Orinoco, Olmstead County, Minn. 

159- Quartz sand, Jordan formation, from Ottawa Sand Co., Ottawa, Minn. 

160. Quartz sand, Jordan formation, at Mazeppa, Goodhue County. 

10 Ries, H., and Rosen, J. A., Report on foundry sands: Ann. Rept. of Geol. Survey of 
Mich., p. 42, 1907. 



ANALYSES OF STEEL SANDS 



65 



Mesh of Screen 



4 — 10 

10 — 20 . . . . 

20 — 40 

40 — 60 

60— 80 

80 — 100 
100 — 200 
200-silt' 

1 5 -minute silt 

30-minute silt 
Clay 



Total 

Comparative mesh 



1 66 A 



.28 

29.14 

62.20 

3.12 

1.52 

1.66 

.12 



98.04 
42 



167A 

.08 

.00 

70.46 

26.00 

1.72 

.58 

.32 

.14 



99-30 
33 



1 68 A 



2.24 
73.82 
22.18 
.56 
.02 
.20 
.16 



99.1; 
3i 



170A 



1.76 
45-95 
35.36 

7.08 

3-22 

3.98 

.38 



97-73 

39 



173A 



5-34 

82.66 

10.48 

•76 

.16 

.02 



99.42 
28 



I75A 



3.64 

58.62 

30.80 

3.96 

1.44 

.48 

• 04 



34 



i8 S B 



1 
39 

38 



99.78 

58 



186B 



15.38 
21.08 
27.94 

15-44 

9.60 

6.30 

•54 



99.38 

55 



166 


Quartz 


sand 


167 


Quartz 


sand 


168 


Quartz 


sand 


170 


Quartz 


sand 


*73 


Quartz 


sand 


175 


Quartz 


sand 


185 


Quartz 


sand 


186 


Quartz 


sand 



Jordan formation, at Kasota, Le Sueur County, Holverson's pit. 

Jordan formation, at Kasota, at River Bridge. 

Jordan formation, i]/ 2 miles west of Merriam Junction, Scott County. 

Jordan formation, at Ottawa, Le Sueur County, Hayes' pit. 

Jordan formation, at Ottawa, Le Sueur County, Rayners' pit. 

Jordan formation, at St. Peter, Nicollet County 

Kettle River formation, at Sandstone, Pine County. 

Kettle River formation, same locality as 185. 



Mesh of Screen 



4 — 10 

10 — 20 

20 — 40 ....... 

40 — 60 

60— 80 

80 — 100 

100 — 200 

200-silt 

1 5 -minute silt 

30-minute silt 
Clay 

Total 

Comparative mesh 



188B 



1.38 

39.08 

22.22 

19.76 

7.48 

5.26 

1.92 

• 34 

1.88 

■ 32 

99.64 

44 



220B 



46.62 

24.62 

14.24 

7-30 

5-92 

• 74 
.00 
.00 

• 50 

99-94 

42 



188. Quartz sand, from Sioux quartzite, Nicollet County, near New Ulm. 
220. Quartz sand, Jordan formation, at Ottawa, Minn., Potters' pit. 

MINERAL ANALYSES OF STEEL SANDS 



Number 

Part examined (over 200 mesh) 

Quartz 

Limonite 

Biotite 

Tourmaline 

Kaolin 

Magnetite 

Chlorite 

Feldspar 

Chromite 

Hornblende 



97-37% 
99.00 

.28 

.01 

.01 



136 



99.1% 
99.2 
.03 



.01 



185 



93 


• 5% 


99 


17 




01 




10 




01 




01 




01 




01 




01 



95.1% 
98.57 
.81 



66 THE FOUNDRY SANDS OF MINNESOTA 

MECHANICAL AND MINERAL ANALYSES OF BRASS SANDS 

The laboratory results of the mechanical analyses of brass sands are 
recorded in Table XL About 75 samples of this type of material were 
examined, and 50 samples were subjected to analysis by both the dry 
method "A" and the wet method "B." The 34 analyses listed in Table 
XI are typical. The dry method analyses are omitted, except No. 70. 

It will be observed from Table XI that the clay content of these sands 
ranges from 2.64 to 19.12 per cent. The 15-minute and 30-minute silts 
combined range from 4.00 to 36.47 per cent, and the comparative mesh 
ranges from 99 to over 250. The mesh is not stated for these sands but 
was calculated and nearly all ran above 150. 

Of the 34 samples listed in Table XI, 18 were collected in the foun- 
dries and represents the type of material in actual use in molding in brass, 
bronze, aluminum, and the alloys of these metals. The remaining 16 
samples were collected in the field in various parts of the state. 

Four of the samples, i.e., 12, 213, 217, and 260, are floor sands; that 
is, sands that have been used one or more times, and consist of mixtures 
of raw brass sand and other materials, such as St. Peter sand and loam. 
The average mesh of these floor sands, which may be taken as represent- 
ing the texture and constitution preferred in this class of work, ranges 
from 132 to 195, the clay content ranges from 3.46 to 5.80 per cent, and 
the silt content ranges from 4.94 to 11.28 per cent. 

Four of the samples, i.e., 18, 210, 221, and 261, are so-called Albany 
sand, supposedly from Albany, New York. But the term Albany sand has 
come to be used as a trade name, and sands are sold under that name that 
do not come from Albany, New York. It is not known with certainty 
that any of the samples listed actually came from Albany. The founders 
only knew that the prices paid, including freight, were enough to have 
warranted the sands coming from that point. 

Number 75 Peter's pit, Fort Snelling. Furnishes brass sand 

Part examined (over 200-mesh) 74% 

Quartz 95 .82 

Hornblende 1.12 

Feldspar 1.08 

Biotite .85 

Muscovite .15 

Augite .12 

Magnetite .10 

Iron oxide .05 

Comparing the so-called Albany sands, Table XI, it will be seen that 
the clay content ranged as follows, 7, 3.74, 4.30, and 10.56 per cent. The 
silt content in these Albany sands ranged from 5.64 to 11.76 per cent. ; 

The two brass sands from Missouri, samples 10 and 29, show the 
same wide variation in both clay and silt content ; that is the clay content 



ANALYSES OF BRASS SANDS 67 

is 11.68 and 3.98 per cent respectively in the two sands, while the pro- 
portion of silt is in reverse order, viz., 7.66 and 21.6 per cent respectively. 

The other brass sands exhibit the same wide range both in clay and 
silt content and in average mesh, from which it is apparent that the 
analyses afford no basis for a standard. 

Sand 260, which was used for extremely heavy casting in brass, car- 
ried 3.96 per cent of clay and 4.94 per cent of silts, and may be close 
to the ideal brass sand. It is a mixture of so-called brass sand and other 
materials; it probably carries a considerable percentage of St. Peter sand. 

Samples 16, 46, 48, and 51 represent material commonly present in 
the glacial drift in the vicinity of Minneapolis and St. Paul, occurring 
as irregular pockets and lenses. In view of the irregularity of the de- 
posits and their lack of uniformity in quality, they are of little importance. 
Samples 75, 76, jy, 82, and 84 were collected in the Peters pit at Fort Snel- 
ling, which was the only pit furnishing brass sand to foundries at that time. 
These samples were collected in different parts of the pit and at different 
horizons from the surface to the floor of the pit, which was about 7 feet 
deep. The series shows the variation that may occur in the character 
of the material in this pit within a distance of a few rods. By selection 
of horizons in certain parts of this pit a very good quality of brass sand 
could be obtained, but uniformity in the grade is hardly to be expected. 
The analyses show that the clay content of the sands in this pit range 
from 2.87 to 13.64 per cent, and that the silts range from 4.08 to 31.23 
per cent. 

Samples in, 124, 208, and 239 represent loess, vast quantities of 
which occur in the southeastern part of the state from Red Wing south- 
ward. The distribution is shown in a very general way in Figure 2. 
Thirty samples of loess were collected in various parts of the area and 
subjected to mechanical analysis. The four samples above mentioned, 
which appear in Table XI, are fairly representative of the group. The 
material is remarkably uniform in physical construction and is essentially 
the same as brass sand. In fact two tons of samples 208 were shipped 
to St. Paul and distributed among four different foundries and used as 
molding sand in brass with entirely satisfactory results. 

Samples 295 and 296 were shipped in from Milwaukee, 295 being 
known under the trademark of "A.A.A." and 296 under the trade 
mark "A." 

Because of the low temperature required for the melting of brass 
and its alloys, mineral composition would be expected to be less im- 
portant than in other sands for foundry use. An attempt was made 
to make a mineral analysis of loess to contrast with that made on the 



68 



THE FOUNDRY SANDS OF MINNESOTA 



St. Peter, but the sand was too fine to give satisfactory results. The 

chemical analysis shows about what minerals are to be expected. See 

page 29. They undoubtedly include considerable feldspar, kaolin, and 
iron oxides. 

XI. MECHANICAL ANALYSES OF BRASS SAND 



10. Brass sand, Gas Traction Foundry, Minneapolis, from St. Louis, Mo. 

11. Brass sand, Union Brass Works, St. Paul, from Monmouth, 111. 

12. Floor sand, brass, Union Brass Works, St. Paul. 

16. Lens of sand in glacial drift, Lawson and Mackubin Streets, St. Paul. 

18. Brass sand, American Hoist & Derrick Co., St. Paul, from Albany, N. Y. 

29. Brass sand, Northern Malleable Iron Works, St. Paul, from St. Louis, Mo. 

30. Brass sand, Northern Malleable Iron Works, St. Paul, from Fort Snelling 
32. Brass sand, Herzog Foundry, St. Paul, local, glacial drift 



Mesh of Screen 


10B 


11B 


12B 


16B 


18B 


29B 


30B 

.12 

• 74 

1. 54 

11.32 

24.50 

49.10 

2.44 

4.86 

5.i6 


32B 


10 — 20 




.04 
.16 
•25 

• 50 
1.42 

18.31 
64.74 

• 57 
3-45 

10.36 


.90 
i.4« 

4.06 

4.84 
22.98 
53-o6 
2.66 
4-32 
3-46 


•32 

•44 
2.08 

•74 

• 50 

2.10 

66.22 

10.48 

14.68 

6.30 


.20 

• 30 
.60 

• 54 
2.32 

3.38 
22.82 
53.48 
1.90 
7.00 
7.00 


.12 
.00 

• 34 
.12 

• 52 
72.20 
11.82 

9.78 
3.98 


.28 


40 — 60 


1.36 

.92 

1 1.52 

.96 

2.52 

60.58 

6.86 


60 — 80 


.01 

.02 

.04 

80.10 

1.94 

5.72 

11.68 


80 — 100 


100 — 200 


15-minute silt 


30-minute silt 


13.44 
9.66 


Clay 




Total 


99.51 


99.80 


97.68 


103.86 


99-54 


98.88 


99.78 


98.10 





46. Pocket of sand in laminated clay, North Minneapolis, glacial. , 

48. Pocket of sand in glacial drift, Lyndale Ave. S. and Minnehaha Drive, 

51. Pocket of sand in drift, Forty-sixth Street S. and Minnehaha Drive, 

70. Brass sand shipped in to Minneapolis, source unknown. 

75. Fort Snelling sand pit, Peters, furnishing brass sand to foundries. 

76. Fort Snelling sand pit, Peters, furnishing brass sand to foundries. 

77. Fort Snelling sand pit, Peters, furnishing brass sand to foundries. 
82. Fort Snelling sand pit, Peters, furnishing brass sand to foundries. 



Mesh of Screen 


46B 


48B 


5iB 

•7i 
.29 
.21 

.17 
.46 

57-74 

19-31 

17.16 

4-50 


70A 


75B 

.22 

1. 71 

5.72 

12.90 

43.83 

26.31 

1.02 

3.06 

5.15 


76B 


77B 

.12 

2.27 

87.83 

4-13 

3-09 
2.87 


82B 


10 — 20 


.08 

.30 

.40 

.30 

.28 

2.14 

84.16 

2.00 

2.00 

8.00 


.08 
•44 
.64 
.40 
.46 
8.04 

70.74 
3.00 
6.00 
9.00 


• 54 

1.84 

4.36 

4.00 

20.28 

68.52 


.21 
1.03 
1.38 
2.35 
18.66 
58.20 
1.03 
8.26 
8.77 




20 — 40 




60 — 80 

80 — 100 


.04 
.19 


100 — 200 


200-silt 


79.22 
2.09 
6.27 
5.96 


15-minute silt 


30-minute silt 


Clay . 




Total 


99.66 


98.80 


100.55 


99-54 


99.92 


99.89 


100.31 


99-79 





Minneapolis. 
Minneapolis. 



ANALYSES OF LOAMS 



69 



Mesh of Screen 


84B 


86B 


111B 


124B 


208B 


210B 


211B 


212B 


213B 


4 — 10 

60— 80 

1 5 -minute silt. . . 

30-minute silt'. . . 

Clay '" . . .' 


2.04 
6.02. 
8.48 
3-36 
2.80 
7.21 
24.77 

7-54 
23.69 
13.64 


.40 

.42 

1. 16 

1-75 

1.68 

i-43 

3.78 
56.65 

6.52 
17.16 

9.04 

99.99 


.06 

.04 

.18 

.20 

1. 16 

59-86 

4.56 

12.00 

19.12 


1. 10 
3.16 
5-94 
3-io 

3-02 

64.74 
2.30 
6.00 
9.00 


.14 
.68 

4-74 

4.18 

4.82 

72.50 

3.84 
5.22 
2.64 


1.30 
1.82 

2.54 
1.88 
8.96 
67.04 
4.68 
7.08 
3-74 


1.56 
1-34 
3-00 
3.28 
9.78 
60.68 
5.96 
9.00 
5-90 


.60 

• 52 

1.64 

1.02 

3.26 

62.64 

7.36 

15.02 

6.14 


,72 
1.98 
1.74 
3.48 
3.96 
8.56 
59.92 
4.10 
7.18 
5.80 


Total 


99-65 


97-i8 


98.36 


98.76 


99.04 


100.50 


98.20 


97-44 



84. Fort Snelling sand pit, Peters, furnishing brass sand to foundries. 

86. Loess, Pilot Knob, near Mendota, upland. 

in. Loess, brass sand, 10 miles northwest of Red Wing, upland, 400 ft. above river. 

124. Loess, brass sand, 1 mile northwest of Nerstrand, Rice County. 

208. Loess, brass sand, Clay Bank station, Goodhue County. 

210. Brass sand, from Albany, N. Y., University of Minnesota, foundry. 

211. Brass sand, St. Paul Brass Founding Co., from "southern Minnesota." 

212. Brass sand, St. Paul Brass Founding Co., obtained locally. 

213. Floor sand, brass, Commutator Company,' Minneapolis. 



Mesh of Screen 


214B 


217B 


22 1 B 


239B 


260B 


261B 


295B 


296B 


297B 


4 — 10 

60— 80 

100 — 200 

1 5 -minute silt. . . 
30-minute silt. . . 


.14 

.00 

.24 

.12 

.22 

65.62 

10.06 

12.92 

9.86 


1.32 
1.44 
5.56 
5-86 
18.26 
56.22 
1.26 

4-34 
4.82 


.12 

•34 

1.54 

.26 

2.82 

38.42 

45-82 

.88 

4.76 

4-30 


.10 

.12 

.18 

.20 

1-54 

75-34 

5.48 

9-90 

4.00 


2.22 

2.84 

2.44 

1.22 

2.2.2 

21.60 

56.92 

.16 

4.78 

3.96 


.32 

.36 

1. 00 

16.34 

60.00 

3-o8 

7.48 

10.56 


.14 

.24 

•34 

4.14 

68.14 

9.30 

13.48 

4.36 


1.42 

12.70 

17.46 

13.29 

30.79 

9.90 

8.80 

1.96 

3.60 


.24 
4.24 

4.84 

4.14 

38.48 

28.09 

2.16 

3-53 

13-44 


Total 


99.18 


99.08 


99.26 


96.86 


98.36 


99.14 


100.14 


99-92 


99.16 



214. Brass sand, Commutator Company, Minneapolis, source unknown. 

217. Floor sand, brass, Twentieth Century Brass Company, Minneapolis. 

221. Brass sand, Commutator Company, Minneapolis, Albany, N. Y. 

239. Loess, brass sand, Farmington, Dakota County, Minn. 

260. Floor sand, brass, Soo Line Railway Shops, Minneapolis. 

261. Brass sand, Soo Line Railway Shops, Albany, N. Y. 

295. Brass sand, from Milwaukee, Flour City Ornamental Iron Company, Minneapolis. 

296. Brass sand, from Milwaukee, Flour City Ornamental Iron Company, Minneapolis. 

297. "French sand," imported from France, Flour City Ornamental Iron Company, Minne- 



apolis; facing sand, ornamental work. 



MECHANICAL AND MINERAL ANALYSES OF LOAMS 

The laboratory results of the mechanical analysis of foundry loams 
are recorded in Table XII. Forty-five samples of loam were collected 
and subjected to mechanical analysis, by both dry and wet methods. The 
thirty-four analyses recorded in Table XII are selected from the list 



;o THE FOUNDRY SANDS OF MINNESOTA 

and are representative. The results by the wet method "B" only are 
given. 

The thirty-four analyses recorded in Table XII include three types 
of material: (a) the raw unused loams taken from the stock bins of 
the foundries, which were used in making facing sands and floor sands, 
and for doctoring floor sands; (b) floor sands in actual use, which consist 
of mixtures of sands, loams, coal screenings (sea coal), etc., and which 
had been used one or more times; and (c) facing sands, which are 
mixtures similar in type to the floor sands, but more carefully prepared, 
and represent approximately, in so far as texture and clay content are 
concerned, the constitution sought in foundry loams. 

The texture of the facing sands, floor sands, and loams is indicated 
in a general way by the "comparative mesh" in Table XII. For the 
facing sands the mesh ranges from 49 to 69, and the clay content ranges 
from 4.56 to 5.64 per cent. For the floor sands the mesh ranges from 
46 to 90, with a clay content of 2.52 to 5.30 per cent. For the raw 
unused loams the mesh ranges from 33 to 250, with a clay content of 4.56 
to 10.98 per cent. 

Some of the floor sands were used without facing sand and others 
with facing sand, and in comparing them the grade of work should be 
taken into account. The most noticeable difference in the floor sands 
for heavy and for light work is seen in the mesh of the sand ; the mesh 
of the sand for heavy work is 62 and 65, in sands 222 and 257, and the 
mesh for light work is 82 and 85 in sands 223 and 258. 

In the 18 samples of fresh, or raw, loam collected in the foundries, 
the analyses of which are recorded in Table XII, the clay content ranged 
from 4.56 to 10.98 per cent. Some of these loams were used directly 
in molding without the addition of other materials, but most of them 
were used for "doctoring" the floor sands ; that is, a little of the raw 
loam was added from time to time to the floor sand to replenish the 
clay as the bond was destroyed by repeated use; and for this purpose 
loam carrying an excess of clay is preferred since a less amount of new 
loam is required to bring the clay content to normal. The silts in these. 
18 samples ranged from 6.10 to 22.J2 per cent. 

All of the loams listed in Table XII were obtained in or near the 
cities of Minneapolis and St. Paul. Most of the loams were furnished 
the foundries by dealers in sands and loams. In Minneapolis practically 
all of the loam was obtained in the northeast part of the city, known 
as Columbia Heights, where the loam occurs as a mantle 2 to 6 feet deep, 
overlying the glacial drift, which forms a succession of small rounded 
knolls with intervening basins or kettle holes, that make the moraine of 
the last glacial period. 



ANALYSES OF LOAMS ' 71 

This loam is subject to wide variations in character within distances 
of a few rods laterally, and varies also from the surface downward, so 
that it is all but impossible to obtain a uniform product by the method 
of pitting it. A uniform product might be obtained by taking large 
batches and machine mixing the batch by the aid of a sand elevator and 
a series of riddles. In the absence of such mixing the loam requires 
close attention and repeated tests to know the actual constitution. 

In St. Paul the loam occurs in the same relation, viz., mantling the 
glacial drift, and is obtained at numerous places, there being no well- 
defined or preferred area. In fact a considerable part of this loam is 
supplied by contractors making excavations for building foundations, 
who deliver to the foundries any material having the general appearance 
of loam that will pass inspection at the foundry. 

In the vicinity of St. Paul and Minneapolis glacial drift of two periods 
is present. The upper or younger drift varies greatly in character from 
place to place, is essentially clayey in constitution, and rarely serviceable 
for foundry purposes. Beneath this young drift is an older deposit known 
as the Early Wisconsin, or red drift, which is less clayey and less variable 
in character. This red drift consists essentially of coarse sand with a 
moderate clay content, and in places very little gravel. A fair example 
of this material is shown in sample 254. Small amounts of this material 
were observed in a few foundries, having been provided by contractors 
as above noted. This material is worthy of a careful test by the foundries, 
since the clay it contains is more refractory than the clays in the surface 
loams, the body of the material is coarser, and in its natural state more 
nearly conforms to the requirements of molding loams. 

Sample 163, Table XIV, was collected at Bellechester, Goodhue 
County, 20 miles by rail south of Red Wing. Extensive pits are operated 
here, mining a semi-refractory clay (Cretaceous) for the Red Wing pot- 
teries. Interbedded with this clay are thin beds of sand which are care- 
fully separated from the clay and piled by means of cable and drag line 
in immense dumps. Sample 163 is a sample of one of the dumps repre- 
senting hundreds of thousands of tons. This material is worthy of a 
careful test in the foundries because: (a) the clay in the sand is as 
refractory as any in the state; (b) the sand is nearly all quartz and 
well proportioned; (c) it is a waste, or by-product of the clay-mining 
and could be loaded on the cars at the same expense that it is piled in 
the dump, since the drag line runs over the gondolas for loading the 
clay; (d) the product would be more nearly uniform than the local loams 
now in use. It will be observed that by analysis 163, Table XIV, com- 
pares favorably with 22, Table XII, which is a synthetic sand and sup- 
posedly is correctly proportioned. 



72 THE FOUNDRY SANDS OF MINNESOTA 

Synthetic sands and loams. — It might seem that synthetic sands offer 
the solution to the problem of foundry loams ; that is, if carefully sized 
quartz sand were mixed with just the right proportion of good clay a 
loam could be made that would exactly meet the molding requirements. 
But when it is recalled that as soon as the loam is used in founding the 
clay is burned to some extent, and with each repeated use the clay bond 
still further deteriorates, it is apparent that no matter how carefully 
prepared originally, the sand rapidly changes in physical constitution 
with use. Not only is the clay destroyed as such, but the burned clay 
and burned sand become dead, silty material that does not function in the 
granulation, but serves instead to clog the interstitial spaces in the sand, 
interfering with the permeability. Synthetic sand therefore deteriorates 
with use, and sooner or later requires doctoring to replenish the bond, 
and to offset the silty material that gradually accumulates with use. A 
carefully proportioned synthetic sand rapidly approaches a constitution 
having the same objectionable features as natural loam. 

The feasibility of synthetic sand resolves itself into a question of 
costs. If natural sands or loams, that approach somewhat closely the 
requirements, can be obtained it is usually cheaper to use these, bringing 
them to the standard by doctoring them at the beginning, rather than to 
build up the entire body by carefully proportioning several sands and clays. 
If natural sands or loams can be obtained the clay content of which is 
refractory, or semi-refractory, with the sand itself also refractory, the 
need for doctoring will be reduced to the minimum, and a long lived sand 
secured. In the light of present foundry practice this seems the more 
likely solution of the problem. 

The clay content of each sample of loam is shown in Table XII. 
This is mechanically separated and not all of it is necessarily the mineral 
kaolinite, though much kaolinite is present. The minerals of the coarser 
portions were determined in five different loams. The quartz content of 
the portions examined ranges from 67 to 90 per cent. Feldspar is the 
most abundant mineral, next to quartz ; and in the coarser sands, frag- 
ments of igneous rocks consisting of two or more minerals make up as 
much as 11.7 per cent of the sand. The figures for iron oxide include 
only grains of that material, and do not include coatings on grains. The 
amount varies from .27 to 1.61 per cent. 

No extensive experiments were carried on to determine the effect 
of mineral composition on bonding power and refractoriness. The fact 
that all the sands except 254, which is a field sand, are being used suc- 
cessfully for some type of foundry work, indicates that aside from the 
clay content, the mineral composition within the limits shown is of 
minor importance. 



ANALYSES OF LOAMS 



TABLE XII. MECHANICAL ANALYSES OF LOAMS 



73 



Mesh of Screen 



4 — 10 

10 — 20 

20 — 40 

40 — 60 

60— 80 

80 — 100 
100 — 200 
200-silt 

1 5 -minute silt 

30-minute silt 
Clay 



Total 

Comparative mesh 



.82 
6.98 

18.74 
7.40 
6.10 
15.46 
32.08 
2.32 
4.44 

5-02 

99.36 

79 



.86 

5-74 

14.10 

6.08 

4.22 

10.98 

38.00 

2.00 

6.00 

10.80 

98.78 



• 54 

5.82 

8.46 

12.90 

6.66 

10.48 

32.26 

4.92 

6.88 

9.96 



.12 

5-44 

15.52 

22.00 

5-30 

9.30 

22.80 

3-34 

5.52 

9-32 

98.66 

85 



17 



3.62 

4.88 

16.32 

12.62 

14.70 

8.44 
12.06 
14.40 
1. 16 
4-94 
6.18 

99.32 

44 



17X 



.70 
2.08 
3.5i 
1-73 
1.64 
7.67 
62.12 
4-39 
9.87 
6.24 

99-95 

214 



.14 

.70 

5-70 

13.78 

24.12 

20.00 

15.00 

4.38 

.38 

5.14 

11.42 

100.76 
70 



23 

.48 

6.52 

15.28 

10.12 

11.22 

6.92 

9.18 

17-50 

1.52 

6.76 



95.62 

52 



2. Floor sand, University of Minnesota, foundry, grey iron work. 

3. Loam, raw, University of Minnesota, foundry, local, glacial. 

6. Loam, raw, Eagle Foundry, Minneapolis, local, glacial, grey iron. 

7. Loam, raw, Gas Traction Foundry, Minneapolis, local, glacial, grey iron. 
17. Loam, raw, Valley Iron Works, St. Paul, grey iron, heavy work, local. 
17X. Loam, raw, Valley Iron Works, St. Paul, grey iron, light work, local. 

22. Synthetic sand; mixture of sands 19, 20, and 21, ratio 1:1:6. 

23. Loam, raw, American Hoist & Derrick Co., St. Paul, grey iron, heavy work. 



Mesh of Screen 



4 — 10 

10 — 20 . . . . 

20 — 40 

40 — 60 .... 

60— 80 

I 80 — 100 
100 — 200 
200-silt 

1 5 -minute silt 

30-minute silt 
Clay 



Total 

Comparative mesh 



24 



.12 
1.60 
1.42 
2.88 
3-o6 
1.32 
9.12 

5-54 
8.40 

5.52 



230 



27 



.60 

1. 00 

4.76 

3-40 

3.38 

1.92 

3-32 

^.28 

7.12 

[5.60 

[0.98 



100.36 

145 



33 



8.40 

5.56 
[2.84 
9.66 
9.78 
6.02 
9.62 
:7.8o 
2.82 
9.02 
8.38 



99.90 
36 



.10 

2.02 

9.64 

12.56 

16.46 

11.56 

14-50 

22.14 

1.56 

4.60 

4.00 

99.14 
62 



223 



1.20 

7.22 
8.96 

13.98 
8.24 

12.32 

35-54 
1.46 
5.10 



96.90 



224 

.06 
.76 
3.60 
1.98 
3.48 
2.22 
4.12 

50.14 

8.02 

14.16 

10.80 

99-34 

210 



225 

-94 
1.70 
7.26 
6.76 

10.52 
6.52 

11.98 

31-44 
3-88 
8.86 

10.00 

99.86 

77 



226 

.24 
.92 
3.08 
2.14 
3.12 
1.96 
3.26 

51.48 
7.62 

14.60 



97-40 
200 



24. Loam, raw, American Hoist & Derrick Co., St. Paul, grey iron, light work. 

27. Loam, raw, Northern Malleable Iron Works, St. Paul, light work. 

33. Loam, raw, Herzog Foundry, St'. Paul, grey iron. 

222. Floor sand, Minneapolis Steel & Machinery Co., grey iron, heavy work. 

223. Floor sand, Minneapolis Steel & Machinery Co., grey iron, light work. 

224. Loam, raw, Minneapolis Steel & Machinery Co., grey iron, local. 

225. Loam, raw, Minneapolis Steel & Machinery Co., grey iron, local. 

226. Loam, raw, Minneapolis Steel & Machinery Co., grey iron, local. 



74 



THE FOUNDRY SANDS OF MINNESOTA 



Mesh of Screen 

4 — 10 

io — 20 

20 — 40 

40 — 60 

60 — 80 

80 — 100 

100 — 200 

200-silt 

1 5 -minute silt. . 

30-minute silt. . 
Clay 

Total 

Comparative 
mesh 



127 



7.06 
20.72 
12.28 
14.16 

6-34 
9.20 
14.40 
2.26 
6.52 
7.70 

100.64 



49 



240 

i.S2 

i-34 

2.98 

5-68 

13.20 

14.74 

27-52 

21.80 

1.70 

3.62 

2.52 

96.62 

68 



241 

.20 

.62 

2.00 

6.32 

22.90 

16.66 

24.88 

18.62 

1.32 

2.56 

3-40 

99-48 

90 



242 

3-70 
1.22 

2-54 

6.78 

14.72 

14.36 

25.86 

21.46 

1.38 

3.38 

4.12 

99-52 

55 



254 



6.08 

11.00 

20.52 

11. 10 

12.94 

6.40 

9.78 

10.52 

2.00 

4-52 

4-56 

99.42 

33 



255 



100.04 
183 



256 



1.58 
6.70 
8.20 
9.40 
7.20 
8.98 

36.00 
5.02 
7.86 

10.98 



101.92 

57 



257 



1.30 

9-54 

22.86 

9-52 

5-92 

11.42 

25-74 

2.38 

5-24 

5.20 

99.12 
65 



• 76 

6.96 

10.26 

16.32 

6.22 

18.76 

27.04 

2.46 

7.08 

5-36 



85 



227. Loam, raw, Minneapolis Steel & 

240. Floor sand, Northern Malleable 

241. Floor sand, Northern Malleable 

242. Floor sand, Northern Malleable 

254. Loam, raw, old red drift, 1349 

255. Loam, raw, Crown Iron Works, 

256. Loam, raw, Crown Iron Works, 

257. Floor sand, Crown Iron Works, 

258. Floor sand, Crown Iron Works, 



Machinery Co., grey iron, St. Paul. 
Iron Works, St. Paul, light. 
Iron Works, St. Paul, light. 
Iron Works, St. Paul, light. 
Central Avenue, Minneapolis, field sample. 

Minneapolis, grey iron work, local. 

Minneapolis, grey iron work, local. 

Minneapolis, grey iron work, heavy. 

Minneapolis, grey iron work, light. 



Mesh of Screen 

4 — ■ 10 

10 — 20 

20 — 40 

40 — 60 

60 — 80 

80 — 100 

100 — 200 

200-silt 

1 5 -minute silt. . 

30-minute silt. . 
Clay 

Total 

Comparative 
mesh 



259 



.46 

10.32 

12.06 

21.58 

12.78 

12.24 

18.10 

2.18 

4-52 

4-56 

98.80 



69 



266 



6.70 

21.60 

21.68 

6.06 

4.06 

9.36 

16.46 

1.50 

7.28 

5-92 

100.62 
49 



267 



8.38 
21.04 
24.06 

5.82 

4.70 
10.48 
15.38 

1.24 

4.64 
4.26 



46 



268 



5.00 

14-34 
23.00 

6-34 

5.08 

11.88 



5.28 
5-30 

99.78 

55 



269 



5-40 
17.10 
27.94 

6.60 

5.20 
H-34 
15.84 

2.18 
.00 

4.76 

96.36 

50 



270 



6.80 

22.30 

23.06 

5.98 

4.06 

8.96 

15-80 

1.44 

5.98 

5-64 



49 



271 



1.30 
2.40 
6.90 

3-47 
2.68 
10.04 
51.90 
3-72 
9.92 



101.17 

175 



272 

.40 
1 .40 
3-88 
5-30 
2.40 
2,56 
9.10 
48.40 
4.96 
8.64 
8.72 

95-76 
142. 



273 

3-54 

4-94 

13-84 

15-70 

5.64 

4.08 

9.88 

22.64 

2.88 

7.28 

8.86 

99.28 
47 



259. 


Facing sand 


266. 


Facing sand 


267. 


Floor sand, 


268. 


Floor sand, 


269. 


Facing sand 


270. 


Facing sand 


271. 


Loam, raw, 


272. 


Loam, raw, 


273- 


Loam, raw, 



, Crown Iron Works, Minneapolis, mixture, loam, sand, coal. 

, American Brake Shoe Co., Minneapolis, mixture, loam, sand, coal. 

American Brake Shoe Co., Minneapolis, backing sand, grey iron. 

American Brake Shoe Co., Minneapolis, light work, grey iron. 

, American Brake Shoe Co., Minneapolis, mixture, loam, sand, coal. 

, American Brake Shoe Co., Minneapolis, mixture, loam, sand, coal. 

American Brake Shoe Co., Minneapolis, grey iron work, local. 

American Brake Shoe Co., Minneapolis, light work, grey iron. 

American Brake Shoe Co., Minneapolis, heavy work, grey iron. 



CLAYS 



MINERAL ANALYSES OF LOAMS 



75 



Number 



Part examined (over 200-mesh) 

Quartz 

Feldspar 

Igneous rock 

Hornblende .... 

Diopside 

Iron oxide 

Chalcedony 

Augite 

Tremolite 

Magnetite 

I Biotite 

Tourmaline 

' Zircon 

I Bronzite 

Chlorite 

1 Serpentine 

! Garnet 

i 1 Dolomite 

' : Slate 



T7 



72.6% 
70.69 

H-95 

1 1. 71 

1.72 

1. 14 

1.02 

• 37 
.27 

• 13 
.09 
.02 
.01 
.01 
.01 



254 



77-83% 
67.91 

15-53 

11. 11 

2.40 

.56 

.91 

■73 
.06 
.01 



.72 
.02 
.01 



272 



26.04% 
83.04 

9.87 

i-43 

2.64 

■ 27 

1-52 

.07 

• 05 
.01 



273 



80.2% 

73-04 

14.18 

6.30 

2,83 



.66 



• 17 



.01 
.01 



• 43 



• 54 



Kerrick 

96% 
90.1 
7.01 



•72 

•47 



.04 
.01 



CLAYS 

Clay is a naturally occurring physical mixture of various products of 
rock-weathering in a very fine state of division, the individual particles 
of which range from about .005 millimeter in diameter to submicro- 
scopic. The chief physical properties of clay are plasticity when wet; 
large absorptive power of water attended by swelling of the mass and 
the evolution of heat ; absorption of gases and coloring matter ; shrinkage 
on drying attended by absorption of heat ; cohesive power in plastic 
state; and great tensile strength when dry. The constituents, as deter- 
mined by chemical analysis, show a preponderance of kaolinite or some 
other hydrous silicates of alumina commonly considered essential. 

In classifying clay according to its uses, Ries 11 cites 47 distinct uses 
made of clay in the industries and arts. The material used for these 
various purposes has come to be called clay, in common parlance, regard- 
less of whether it is a pure mineral, or consists largely of impurities or 
inert material. In fact kaolinite rarely occurs in nature in the pure state ; 
it nearly always occurs mixed with inert material or impurities such as 
sand and other rock fragments finer than sand, commonly called silts. 
Accordingly we have brick clays, potters' clay, china clays, modeling 
clays, etc. 

Pure clay as such could not be used in most of the industries in 
which so-called clay is employed. Brick clays may carry less than 5 
per cent of kaolinite, the remaining 95 per cent being inert material, 



11 Ries, H., Economic geology, p. 182, 1916. 



7 6 THE FOUNDRY SANDS OF MINNESOTA 

principally sand and silt. Sample 59, Table XIII, shows by mechanical 
analysis only 1.5 per cent of the finest grained material, yet it is used 
for making brick. 

The materials used in ceramic work under the name of clay commonly 
contain less than 50 per cent of kaolinite, but the other inert material j 
present is just as essential as the true clay, so that it is not strange that 
the term clay is used to designate the heterogeneous mixture. By common 
consent any material that will develop a certain degree of plasticity is 
called clay in most of the industries using clay. 

Colloidal material in clays. — It has long been known that certain sub- 
stances in solution will diffuse through membranes whereas other sub- 
stances will not, also that the same substance may be in such a state that 
it will diffuse through a membrane or it may be in such a state that it 
will not. Substances in the solutions that will not diffuse through a 
membrane are said to be in a colloidal state. Some substances exist under 
certain conditions in the colloidal and under other conditions in the non- 
colloidal state. 

Colloidal materials may be organic or inorganic. The organic colloids 
include such common materials as jellies, albumen, gelatin, agar, and 
compounds of humic acid. The inorganic colloids include such common 
substances as silicic acid, ferric hydrate, and aluminum silicate. The 
colloidal state may be simply a matter of the fineness of division of the 
particles. It is thought that any substance can be brought into the col- 
loidal state if it can be reduced to fine enough individual particles. 

It is known that most, if not all, clays contain colloidal material, but 
the per cent of colloidal material in clay is small, rarely more than 3 per 
cent, and usually much less. The colloidal material in clay may be silica, 
alumina, ferric iron, or other inorganic material, or it may include organic 
material also. 

It is thought by some investigators that the plasticity of clays is 
dependent on the colloidal material present in the clay, and that the 
absorptive power of clay for water and for coloring matter is also de- 
pendent on the presence of colloidal material. Moore, Fry, and Middle- 
ton 12 found that true clay, as well as clay soils, absorb coloring matter 
in direct proportion to the amount of colloidal material present. 

Plasticity of clays. — Plasticity is that property in a material which 
enables it to change form without rupture, the new shape being retained 
when the deformitory force is removed. The cause of plasticity has not 
been satisfactorily explained. Some clays are much more plastic than 
others. Some clays that are of low plasticity develop greater plasticity 
under certain kinds of treatment, such as grinding, or weathering. 

12 Moore, C. J., Fry, W. H., and Middlet'on, H. E. Methods of determining colloidal matter 
in soils: Journal of Industrial and Engineering Chemistry. Vol. 13, 1921. 



CLAYS 77 

Clays that are richest in kaolinite are seldom as plastic as those that 

re not so pure. Plasticity does not appear to be connected with chemical 

.imposition, yet on heating to temperatures of 41 5 to 6oo° C. all clays 

|| >se their plasticity, and it can not be restored. It has been urged that 

lasticity is due to the shape of the clay particles, and there is some evi- 

! < ence in support of this claim, but the hypothesis remains unproved. 

Elasticity has been increased by bacterial inoculation, but it has not been 

>roved that bacterial action is the cause of plasticity. 

Whatever its nature and origin, it is fairly well established that many 
if the properties of clay are closely connected with the colloidal matter 
>resent, such matter being in the form of a film of colloidal gel surround- 
ng particles of non-plastic inert nature. 

Tensile strength of clays. — Tensile strength, or the resistance to tor- 
sional stresses, or to rupture in clays is influenced by the plasticity of the 
day, the size and shape of the inert non-plastic grains in the mass, and 
<:he amount of colloidal material present. 

In the mechanical analysis of molding sands, loams, and foundry 
days, the material that would remain in suspension in distilled water 24 
hours after deflocculation with ammoniated water, was called clay. No 
attempt was made to determine the amount of colloidal material present 
in this clay. The tensile strength developed in the molding sand, as 
shown by the briquet tests, varies with the proportion of clay present, 
but is not in direct ratio, that is, the same per cent of clay gives greater 
1 tensile strength in some sands than in others. In the same type of ma- 
terial, that is, material of the same origin, or from the same rource, the 
tensile strength is approximately in proportion to the per cent of clay 
present. 

The variation in the tensile strength may be due to the variation in 

I the amount of colloidal material present in the different samples, or it 
imay be due to differences in the sand, that is the proportioning of the fine 

and coarse sizes, the shape of the grains, etc. On the whole it seems 

II more probable that the difference in tensile strength is due to differences 
in the clay itself in the different samples, which in turn suggests that 

1 the variation in the amount of colloidal material present in the different 
s samples is the real cause of the difference in tensile strength. Much more 
1 detailed experiments would be required to obtain data on which a conclu- 
5 sion might be based. 

Mechanical analysis of clays. — The results of the mechanical analysis 

1 of the so-called clays are recorded in Table XIII. Of the 17 analyses 

I there recorded, 4 are analyses of clays collected in tha foundries and 13 

were samples collected in the field in different parts of the state, and used 

as clay for manufacturing purposes, or were material suitable for such use. 



7 8 THE FOUNDRY SANDS OF MINNESOTA 

The clays collected in the foundries were used for various purposes, 
such as daubing cupolas, patching the lining of steel furnaces, doctoring 
burned out molding sand, and for making synthetic molding sand. The 
content of clay in the 17 samples as shown by Table XIII, ranges from 
1.5 to 69 per cent. Comparing Table XIII with Tables IX, X, XI, XII, 
and XIV, it will be observed that many of the so-called sands and loam* 
carry a larger per cent of "clay" than some of the samples listed in Table 
XIII, as clay. Even the St. Peter sand carries a larger per cent of "clay" 
than sample 59, Table XIII, which is a brick "clay." 

The value of so-called clays for certain foundry purposes is roughly 
proportional to the per cent of clay present. Material that shows 20 per 
cent of clay is worth about twice as much as material that shows 10 per 
cent of clay, and the price paid for the material should be graduated 
accordingly. 

Of the 17 clay samples listed in Table XIII, 8 came from the Cretaceous 
formation either directly or indirectly. They are numbers 10, 20, 129, 
130, 132, 150, 164, and 171. They show a content of clay ranging from 
28 to 69 per cent. Sample 265, supposed to be of more than ordinary 
refractoriness, showed a clay content of 37.02 per cent. This clay was 
imported from some other state, its original source was not learned. 

Sample 164 was collected at Bellechester, Goodhue County, Min- 
nesota, from the pits where clay was being mined for the Red Wing 
potteries. Several grades of clay are mined in this locality, but samples 
were not taken of all. Sample 164 does not fairly represent all the 
clays here found. 

Sample 150, which had the highest proportion of clay, was collected 
a few miles north of Rochester, Olmstead County, Minnesota, from the 
glacial drift. A small mass of Cretaceous clay was here found in the 
drift, having been picked up by the glacier and incorporated without 
mixing with the foreign material. It is of no commercial value in itself, 
but is of interest in showing that deposits of exceptionally good clay occur 
in the region over which the glacier passed. It is within the range of 
possibilities that a systematic search might locate them. 

Sample 130 was collected 2 miles east of Austin, Minnesota, 8 miles' 
north of the Iowa line, from an outcrop of the Cretaceous. 

Samples 45, 53, 58, and 59 were collected in North Minneapolis, in 
the Mississippi Valley above the falls, to illustrate the type of clays that 
occur here in considerable areas. It is improbable that this clay would 
be of much service in foundry work, since it carries a large percentage of 
lime, and other fluxes. 

Clays similar in character occur at numerous points in the valley oi 
the Minnesota River from Mankato northward to St. Paul, a distance 



CLAYS 



79 



of ioo miles. The type is illustrated by sample 180, which was collected 
at Chaska, Scott County, where it is extensively used in the manufacture 
of common brick. 

In a previous investigation of Minnesota clays, 13 samples were col- 
lected and analyzed from the same localities as those of interest in foundry 
work. They serve to indicate in a general way at least, the probable 
composition of the clays listed in Table XIIL 

The clays used in the Red Wing potteries 14 came from Bellechester 
and Clay Bank, Goodhue County, and may be taken as probably similar 
to, if not the same as, 164 given in Table XIIL Three analyses of these 
clays show the sodium oxide, magnesia, lime, and iron oxides, amounting 
to 5.59 to 6.04 per cent. It is probable that samples 10, 20, 129, 130, 132, 
and 150 would show upon analysis about as much of these fluxing con- 
stituents. Similarly, a sample of clay from North Minneapolis '•hows 
18.73 P er cen t °^ potash, soda, lime, magnesia, and iron oxides. This 
is from the same locality as samples 58 and 59, and is probably the same 
type of material. It is probable that samples 45, 55, and 180 would show 
upon analysis a similar composition. From these analyses it will be seen 
that the glacial clays carry more than three times as much of the alkalies, 
iron oxides, and magnesia as do the Cretaceous clays, and they are probably 
correspondingly less refractory. 



TABLE XIII. MECHANICAL ANALYSES OF CLAYS 



Mesh of Screen 


10X 


20 


45 


53 


58 


59 

.14 
4-52 
84.00 
2.00 
5.00 
1.50 


129 


10 — - 20 


.26 
•44 

• 32 

• 50 
.28 
.80 

18.72 

1.62 

19.98 

54.00 


.90 

2.00 

3.60 

2.22 

1.80 

3.46 

12.38 

2.76 

22.00 

48.26 


1.04 
2.. 08 

.56 

57.20 

10.40 

22.88 

5.82 


.20 
.18 
1. 16 
84.98 
3-24 
4-32 
5-42 

99-50 


• 74 

•94 

.24 

.24 

.16 

.62 

20.20 

5-QO 

47.00 

24.00 


1.32 

2.52 


20 — 40 


40 — ■ 60 


1.94 
.70 


60 — 80 


80 — -IOO 


.58 


100 — 200 . 


i-34 
9.10 

2.58 


1 5 -minute silt 


30-minute silt 


22.24 


Clav 


58.26 






Total 


96.92 


99.38 


99-98 


99.14 


97.16 


100.58 







10. Gas Traction Founding Co., Minneapolis, used in steel work, furnace, came from 
1 Ottawa, Minn. Cretaceous, furnace lining, and molding. 

20. American Hoist & Derrick Co., St. Paul, came from Ottawa, Minn. Cretaceous, used 
i in molding sand, and for furnace work. 

45. River clay, North Minneapolis, City Work House, clay pit. 

53. Glacial clay, Minnehaha Creek and Lyndale Avenue, Minneapolis 

58. Marshall Avenue and Twenty-ninth Street N., Minneapolis, six feet below street. 

59. Marshall Avenue and Thirty-fourth Street N., Minneapolis, brick yard. 
129. Ottawa, Minn., sample furnished by Ottawa Sand Co. 

13 Grout, F. F. Bulletin, Minnesota Geological Survey, No. n, 1914. 

14 Analyses by Grout. Op. cit. 



8o 



THE FOUNDRY SANDS OF MINNESOTA 



TABLE XIII. MECHANICAL ANALYSES OF CLAYS— Continued 



Mesh of Screen 



4 — 10 

10 20 

jo — 40 . . . . 
40 — 60 . . . . 

60— 80 

80 — 100 

100 200 . . . . 

200-silt 

1 5 -minute silt' 
30-minute silt 

Clay 



Total 



130 


13-2 


150 


164 


i/i 


180 


.18 












.18 


.60 




.26 






.14 


1.72 




• 54 


.24 




.16 


.86 


.18 


.14 


• 34 




.28 


1. 16 


•32 


\ 1 


.18 




•37 


1.70 


• 54 


• 50 


.20 


50.20 


1.88 


10.10 


.90 


7 -7 2 


•44 




11.86 


23.06 


1.48 


32.38 


11.88 




3-9^ 


.98 


7.70 


4.60 


4.12 


11.28 


12.00 


16.00 


21.24 


23.06 


35-68 


18.34 


67.14 


36.02 


69.32 


27.96 
97.28 


45.08 
98.16 


20.68 


98.15 


92.20 


101.68 


100.50 



253 



1.44 
4.52 
8.24 

3-52 

2.36 

6.16 

36.22 

6.28 

13-50 

17.46 

100.18 



130. Cretaceous outcrop, two miles east of Austin, Minn. 

132. Cretaceous outcrop, bank of Iowa River, one mile above Le Roy. 

150. Lens of clay in glacial drift, near Rochester, derived from Cretaceous. 

164. Bellechest'er, Goodhue County, Minn., clay pit of Red Wing, potteries, Cretaceous. 

171. Ottawa, Minn, clay pits, Cretaceous clay in pot holes in surface of Jordan formation. 

180. Chaska, Carver County, Minn., Minnesota River Valley clay used for common brick, 
glacial. 

253. Eagle Foundry, Minneapolis, local clay loam, glacial, used in daubing cupola, locality 
unknown. 



Mesh of Screen 



4 — 10 . . . . 

10 — 20 . . . . 

20 — 40 . . . . 

40^ 60 . . . . 

60 — 80 

80 — 100 . . . . 
100 — 200 . . . . 
200-silt 

1 5 -minute silt 

30-minute silt 
Clay 



Total 



265 


274 


1.03 




.62 




1. 10 


•50 


2.84 


2-44 


1.44 


2.20 


1.44 


1.84 


4.84 


7.10 


10.74 


41.80 


5.64 


4.38 


28.36 


9.44 


37.02 


28.38 
98.08 


95-07 



275 



.48 

2-60 

9.64 

5-44 
4.40 
12.48 
46.40 
3.46 
6.58 
7.98 

99.46 



265. American Brake Shoe Company, Minneapolis, imported clay, semi-refractory, used for 
furnace and molding. 

274. Columbia Heights, Minneapolis, clay loam, furnished to foundries, extra heavy. 

275. Columbia Heights, Minneapolis, ordinary loam, as furnished to foundries. 



Samples 274 and 275 are included in Table XIII for purposes of com- 
parison. These samples were collected in one of the pits at Columbia 
Heights, Minneapolis, which furnish loam to the foundries. Sample 274 
is a fair average of the type of material here found. This sample is from 
a lenslike horizon of clay loam occurring in the main bed. It will be 
observed that 275 carries more clay than the glacial or river clays above 
described, i.e., 45, 53, and 59, and at the same time is not rated as clay — 
illustrating the confusion of terms. 



SANDS 



TABLE XIV. MISCELLANEOUS SANDS 



Mesh of Screen 



4 — 10 
10 — 20 

20 ■ 40 . . . . 

40 — 60 

60— 80 

80 — 100 . . . . 

100 200 . . . . 

200-silt 

15-minute silt 
30-minute silt 

Clay 

Total . . 



4B 



.61 

.96 

7.78 

23-74 

5.58 

6.02 

13.06 

28.94 

1.88 

4.68 

4.96 



14B 



4-36 

I3-30 

15.16 

14.36 

37.22 

9-90 

1. 00 

2.00 

3-00 

100.98 



25A 

26.00 
29.50 
31.00 
7.80 
2.80 
1. 00 
1.60 

I.IO 



100.80 



61 



.80 
10.06 

22.24. 

32.54 
14.44 

8.73 

3-20 

•50 

2.00 
4.00 

98.51 



97 


139 


141 

.40 

-56 

.82 

.66 

5-70 

50.00 

12.00 

14.00 

I5-I4 


158 


1.50 
75-20 
5.00 
8.00 
8.00 


.80 

.36 

31-88 

51.70 

.18 

4.00 

8.80 


.96 

I.IO 

4.22 

8.76 

30.00 

46.00 

Trace 

4.00 

2.50 


97-70 


97.72 


99-28 


97-54 



4 
14 
25 
61 

97 
139 
141 
158 



Floor sand, Eagle Foundry, Minneapolis. 

Wind blown glacial sand, East Hennepin Avenue, Minneapolis. 

Blast sand, glacial, American Hoist & Derrick Co., St. Paul. 

Dune sand, Fridley, Anoka County. 

Dolomitic sand, 10 miles east of Cannon Falls. 

Dolomitic sand, Fillmore County, 4 miles southwest, W>koff. 

Dolomitic sand, Spring Valley, Fillmore County. 

Dresbach sand, Red Wing. 



Mesh of Screen 



4 — 10 

10 — ■ 20 . . . . 

20 — ■ 40 .... 

40 — 60 . . . . 

60 — 80 

80 — 100 
100 — 200 . . . . 
200-silt 

15-minute silt 

30-minute silt 
Clay 



Total 



161A 

18.24 

32.50 

27.20 

9.84 

2-54 
1.96 
3-30 
3-86 



99-44 



162A 



1.78 

20.12 

43-92 

16.90 

7-56 

5.06 

1.96 



98.26 



163B 



3. 
30. 
18. 
26, 

9 



8.76 



101.26 



I78B 














10 

84 

I 


10 
08 
22 
50 
So 
62 
80 


97 


82 



i8ib 



.04 

.16 

21.14 

74.96 

2.24 

•52 

1.02 



200B 



71-30 
25.28 
.98 
.58 
.08 
.10 
.16 
.00 

.48 
.62 

99-58 



205B 



■30 

8.80 

15-90 

25-92 

16.36 

16.66 

9.76 

1.24 

2.98 

1.56 

99.48 



ioyB 



98.66 



161. Cretaceous sand, Bellechester, Goodhue County. 

162. Cretaceous sand, Bellechester, Goodhue County. 

163. Mixed sands in dump, Bellechester, Goodhue County. 
178. Silty River sand, Chaska, Carver County. 

181. Silty River sand, Carver, Carver County. 

200. Quartzite screenings-, Jasper, Rock County, from crusher. 

205. Quartzite screenings, Jasper, Rock County, from crusher. 

207. Quartzite screenings, Jasper, Rock County, from crusher. 



136343 



82 



THE FOUNDRY SANDS OF MINNESOTA 



MINERAL ANALYSIS OF MISCELLANEOUS SANDS 



Number 



Part' examined (over 200-mesh) 

Quartz 

Feldspar 

Limonite 

Hornblende 

Kaolin 

Chlorite 

Diopside 

Magnetite 

Rock 

Zircon 

Serpentine 

Tourmaline 

Spinel 

Garnet 

Mica 

Glauconite 



61 



66% 

05 

5'i 



489 
192 
072 
042 
038 
008 
005 
004 
002 
001 
001 



40.7% 
98.841 
.26 
.003 



.016 



•49 
.006 



163 



97-5% 
96.54 
2.96 
.01 
.009 



LABORATORY TESTS OF TENSILE STRENGTH 

The laboratory work was guided to a large degree by the sand troubles 
encountered in the foundries visited, and was accordingly devoted to 
specific problems for the most part, rather than to systematic series. 

Samples B 4, C 4, and D 4, all in Table XV, were collected in the 
same foundry, but from different sand piles. In this foundry the prac- 
tice was to allow each molder to follow his own method of tempering 
the sand, and the purpose of the tests was to determine to what degree 
the personal equation of the molder was a factor in the tensile strength. 
The same raw sand was used by each molder. It will be observed that 
there was a variation of 50 per cent in the moisture content, of 15 per 
cent in the clay content, and a considerable variation in the tensile 
strength, in these three samples. The tensile strength in these three 
samples is not in proportion to the content of clay or of moisture. The 
maximum strength coincides with the minimum moisture and a medium 
clay content. Inasmuch as the same raw sand was used throughout the 
foundry, the variation in the clay content in the different piles is 
plainly due to doctoring the sands ; that is, to the amount of raw 
sand added from time to time to replenish the bond as it is burned out, 
the amount so added varying with the individual molder. 

Samples 240, 241, and 242 were collected in another foundry, but in 
the same manner as above described ; that is, from separate heaps of sand 
in actual use for the same grade of work, and made by mixing the same 
original raw sands. In this instance, however, the sand for the entire 
foundry was tempered and riddled by one man whose business it was after 
closing hours, during the night, to prepare the sand for the following 
day. Even in this case, where exceptional care was exercised to obtain 



LABORATORY TESTS OF TENSILE STRENGTH 83 

uniformity in the sand, at least in so far as riddling and tempering the sand 
was concerned, there was still a considerable variation in the moisture, 
in the clay content, and in the tensile stength, indicating the need of some 
simple and easy method of making accurate determinations of these factors 
from day to day. The problem in this instance was not regarded as 
userious, but it was recognized by the foundry management that there was 
a lack of uniformity in the grade of castings turned out, which was 
attributed to the sand used, and that the difficulty might be corrected if its 
nature could be discovered. 

Samples 222 to 227, inclusive, were collected from still another foundry 
where the sand trouble presented a problem of a more serious nature, 
because of the loss from defective castings. In this instance sands 224, 
225, 226, and 227 represent the raw sands, or loams, furnished the foundry. 
^Samples 222 and 223 were floor sands in actual use, taken from the sep- 
arate sand heaps, and molders' tables, and were made by mixing the loams 
:224, 225, 226, and 22*] in varying proportions. 

The sand did not stand up in the mold, in consequence of which blisters 

land blow holes appeared in the castings. The mechanical analysis of 

sjsand 222 showed a clay content of 4 per cent which should have afforded 

lample bond. The moisture, however, was abnormally low, viz., 3 per 

cent. The tensile strength was so low as to be negligible. Adding water 

in the laboratory up to 10 per cent in this sample failed to develop the 

required strength. The tensile strength of the raw sands 224, 225, 226, 

and 227, as the table shows, was ample. This seemed to indicate a burned 

iout sand, and a failure to add enough raw sand to counteract the burning 

from repeated use. However, the presence of 4 per cent of clay in 

sample 222, according to the analysis, Table XII, does not appear to be 

tin harmony with this suggestion. 

A second visit to this foundry showed that the work here consisted 
in making large, rather massive, castings in grey iron and semisteel. 
It was necessary for the castings to remain in the flask for a considerable 
time before being shaken out, and the sand of the entire mold became 
1 heated excessively. Without allowing the sand to cool down to normal, 
>or to rest, water was added to replenish that lost from evaporation and the 
^operation was repeated. Under this treatment the clay which should 
«erve as the bond had no opportunity to function. 

Time is an important factor in developing plasticity in clay, which 
^appears to have been disregarded in this case. The trouble might be 
inscribed to what may be called a fatigue of the bond. The most obvious 
remedy was the use of a larger volume of sand so that the sand might 
lhave a rest between each period of use, allowing the water to be absorbed 
'by the clay. 



84 THE FOUNDRY SANDS OF MINNESOTA 

Samples 200, 201, 205, and 207 give the results of an experiment 
with crushed quartzite to determine the effect of extreme angularity of 
grain on tensile strength. From the mechanical analysis of these samples, 
Table XIV, it will be seen that 200 is coarse, 205 medium, and 207 fine 
sand. Number 200, which showed a clay content of .62 per cent, did not 
develop a measurable amount of tensile strength. It is of interest to 
compare this sand with sand 5 from the St. Peter formation, in which the 
clay content is even less, but the sand grains are well rounded, and in 
which a tensile strength of 2.8 pounds was developed in the air-dried 
briquet. 

The individual grains of quartzite screenings represent the extreme 
of angularity, a large proportion of the grains being long, sharp needles 
or elongated prisms. The experiment, while not elaborate enough to 
carry much weight, points to extreme angularity of grain as a disadvan- 
tage rather than an advantage in tensile strength. 

A second experiment was tried to determine to what degree the 
tensile strength test could be depended on to reflect the varying content 
of bond. For this purpose a series of mixtures of sand 1 and sand 208 
were used, ranging from 90 per cent of the former, and 10 per cent of 
the latter, to 10 per cent of the former and 90 per cent of the latter. 
The results are recorded in Table XV, in samples 229 to 237 inclusive. 
The clay content and tensile strength of samples 1 and 208, the components 
of the mixtures, are shown elsewhere in the table. It will be observed 
that the clay content is comparable in the two, but that the tensile strength 
is twelve times as great in 208 as it is in 1. 

From Table XV it will be observed that the tensi 1 e strength increases 
quite uniformly with the increase of the proportion of sand 208 (loess) 
until 80 per cent of this sand is reached, after which the change is not 
appreciable. 

DETERIORATION OF SAND WITH HEAT 

It is well recognized that the bonding power of sand deteriorates with 
repeated use, and unless the sand is doctored by adding new raw sand, 
or clay, or some artificial bond, it soon becomes so weak that it will not 
stand up in the mold. In this condition it is said to be burned out. 

When the liquid metal is poured into the mold the sand on the face . 
of the mold is heated to temperatures approaching the temperature of the 
melt, which in brass and its alloys is 800 to 900 degrees F. and in cast steel 
is about 3,600 degrees F. Back from the face of the mold the sand is 
heated to progressively lower temperatures. But while the sand in the body 
of the mold is heated to moderate temperatures only, it nevertheless de- 
teriorates in strength. By way of demonstrating this the following 
experiment was undertaken. 



DETERIORATION OF SAND WITH HEAT 85 

Three samples of loess were taken from samples 208 and labeled 283, 
284, and 285. They were placed in an oven at room temperature and 
the oven raised to a temperature of 400 degrees C, or 752 degrees F., which 
temperature was maintained. At the end of one hour sample 285 was 
removed. At the end of two hours sample 284 was removed, and sample 
283 was allowed to remain four hours. After baking, the samples were 
subjected to the wet method mechanical analysis to determine the amount 
of clay that would stay in suspension in water. Briquets were made 
from the material after tempering with water, and the briquets tested 
for tensile strength with the results shown in Table XV. 

The clay content of the original raw material, loess 208, as shown above, 
is 2.64 per cent, and the tensile strength is 18.2 ounces. The tensile 
strength decreased in proportion to the length of time the material was 
baked, the decrease amounting to nearly one third in sample 283 after four 
hours baking. The clay content decreased also, amounting to nearly 35 
per cent in sample 283, baked four hours ; and less amounts for the other 
samples baked shorter periods. 

A second experiment of the same nature was undertaken using two 
surface loams from Columbia Heights, viz., 274 and 275, renumbering 
them 286 and 287. Both samples were baked four hours at a temperature 
of 400 C, after which the clay content was determined by the usual 
method, and the tensile strength determined by briquets. The mechanical 
analysis of samples 274 and 275 are shown in Table XIII, from which 
it will be seen that sample 274 before baking showed a clay content of 
28.38 per cent, whereas after baking four hours, as sample 286, Table 
XV, it showed a clay content of only 5.84 per cent, or a loss of 79 per 
cent. 15 

Sample 275, as will be seen from Table XIII, showed a clay content of 
7.98 per cent, whereas after baking four hours, as sample 287, Table XV, 
it showed a clay content of only 4.26 per cent, or a loss of 46 per cent. 

The effect of baking these loams was probably partially to dehydrate 
the clays, in consequence of which they lost some of their plasticity and 
were less effective as bond. The clay also lost in part the property of 
deflocculating when churned in ammoniated water, and hence would not 
stay in suspension. 

The results of these experiments indicate that the mechanical method 
of determining the clay by suspension in ammoniated water gives an 
approximate measure at least of the amount of clay present that is effec- 
tive as bond, and that the tensile strength test by means of briquets is at 
least an approximate measure of the amount of such clay present in the 
sample. 

15 The term clay here used refers to the material that would stay in suspension in water 
24 hours. 



86 



THE FOUNDRY SANDS OF MINNESOTA 



a 

o 
O 



























>*> 






























































79 










d 


























P^ 










d 


























03 










o 


























In 




09 


-d 

09 




U 


























d 




T3 


X3- 




d 


























o 




Td 


•d 




























d 




03 


o3 




















>* 


-* 








d 










Td 












a 




PQ 


W 


a 






a 


<u 


!>> 


!>> 




o 
o 
O 












o 








o 






u 


03 


o3 


03 




bo 










a 




en 


03 


a 






o 


hi 
03 
O 
19 


o 

a 


O 




_d 

•d 
d 










u 








hi 
















o 


d 
o 

> 

03 

09 
J3 

03~ 
03 

d 










o 

u 

<u 
+-> 
19 

Ph 

a 


;>, 

u 
-d 
d 
d 

o 

M-l 

a 


d 

Pi 
o 

a 

o3 
to 

a 


X) 

a 

J3 
O 

<-hl 

19 

a 

03 

a 


o 

hi 

<u 
Ph 

a 


•d 

09 

id 

in 


<u 

X) 
03 

!>> 
13 
in 


03 

a 

hi 

o 

M-l 

hi 
U 

i3 
Ph 


a 

13 

a 

03 
bo 

.a 


o 

<L9 
Oh 

Pq 

tn 

bo 

'3 


CJ 

hi 

09 

a. 
pq 

of 
bo 
d 

'S 


d 

hi 

09 
t> 

bo 


-h> 

o3 

09 

C 
o3 

pq 

03 


>> 

hi 

-d 
d 
d 
o 

.a 

-d 

09 


hi 

-d 
d 
d 
o 

_d 
*d 

19 


d 
d 
o 

4H 

_d 

19 




o 


o 


o 












09 


OJ 


U 












o 
u 


hi 


u 


hi 


u 


09 

03 


09 


w 


03 

hi 


09 
U 


09 
hi 


09 

hi 


d 


d 


d 


d 




d 


t3 


d 


-d 


-d 


-d 


-d 


a 

o 


19 

03 

09 


o 

09 

19 


O 
09 

19 


19 

03 

09 


a 

o 
u 


d 


03 


03 


PS 

03 




03 


c3 


oj 




03 


03 












«+-i 


^ 














en 


03 




















09 










in 








m 




























u 

o 


in 

o 
o 


O 
O 


hi 

o 

O 


hi 
O 


13 

a 


a 

03 


x3 
el 

03 


U 
03 

d 


U 

d 


hi 

o3 

d 


1-1 
03 

d 


03 
09 
09 
O 


Sh 
O 

o 


a 

o 


a 

O 


a 

03 

O 




U 


to 


h 


ft 


U 


C/3 


C/2 


C/3 


a 


a 


a 


a 


h-1 


h 


h-1 


h-1 


h-1 





PQ 



pq 



09 

-d 
d 
d 

o 
Ph 

d 

h- 1 




00 


0\ 






















M 


O 
in 


t^ 

t^ 


O 

6 

oo 





<; 



0) 






































'd 






















VO 
















d 
d 

o 
Ph 




hi 


oo 


cq 


oo 


O 


oo 


o 


o 


vd 


oo 


o 




oq 


CS1 


o 


■* 




a 


LO 


6 


OO 


<N 


lO 


t>. 


VO 


M 


M 


M 


M 






pq 


m 


oo 








M 










0) 








M 






CM 


H 


M 




d 






































h- 1 







































o 



09 






































09 






































19 






































d 


lO 


VD 


oo 


in 




■* 


I- 1 


in 


O 


<-o 


o 


CO 


09 


in 


O 








d 


w 


O 


In. 


cy, 


H 


CO 


H 


t 


C9 


CS 


pq 


■* 


00 




o 








U 




H 












H 








H 


M 




M 








d 






































i— i 







































H « H 
>.ZHJJ 
< W S U 
J H W m 

Ph 



\r> 


•o 


0) 


uo 


in 




in 


^o 


pq 


pq 


VO 


oo 


^h 








00 




M 


Oi 


hi 


^ 


O 




o 


M 


\© 


MD 


in 


00 


vo 


o 


oo 


O 


0\ 


In. 


oq 


■tf 


in 


4- 






0) 


pq' 




pq 


co 


CO 


pq 


-t 


o 

hi 


o 


00 


IN. 



P3 



« £ H 

& S z 

H W W 

o 5 «: 







in 


in 






























in 


tN 


o 


tN 


in 


in 


in 


in 


m 


m 


in 


M 




CO 


VO 


VO 


VO 


*o 



< 


w 


tM 


« 


pq 


>■ 


o 


^ 


< 


<; 




J. 


hJ£ 







■* 


Tt" 


xt 


in 


< 
in 


PP 
in 


O 


o 




in 


In. 


00 


pq 


■sh 


in 


vo 


In. 




PP 


u 


n 




C\ 


o 
pq 


o 
pq 


o 
pq 


o 

01 


o 
pq 


pq 
pq 


pi 
pq 


pq 
pq 


pi 
pi 


pq 
pi 



STRENGTH OF SAND BRIQUETS 



87 



H 



U 



P 
H 

< 



CO 


00 


CO 


00 


CO 


00 


co 


no 











O 











O 


PI 


PI 


p] 


PI 


Pq 


pq 


0) 


<N 


T3 


X) 


X5 


tJ 


T) 


Tj 1 


Tf 


•n 


£ 


cj 


rj 


C 


a 


3 


a 


3 


rt 


a 


rt 


a! 


aj 


rt 


rt 


crt 


co 


CO 


co 


CO 


CO 




CO 


CO 


+-> 


4_> 


^j 


-*-> 




+j 


4J 




£ 


s 


£ 


£ 


£ 


3 


3 


£ 




cu 


V 


cu 


<u 






cu 


CJ 


cj 


CJ 










U 


CJ 


Jh 


Sh 


i-l 


In 


tH 


j-. 


Vh 


!h 


CU 


cu 


CJ 


CU 


CU 




CU 


cu 


ft 


ft 


ft 


ft 


ft 


ft 


ft 


ft 


O 














O 








M 


PO 


•* 


in 


VO 


i^ 


CO 


On 


M 


H 


M 


M 


M 


M 


M 


M 


T) 


T3 


-tf 


T3 


-d 


TJ 


TJ 


T) 


c 


c 


£ 


£ 


a 


3 




£ 


oj 


nJ 


rt 


rt 


cd 


a3 


rt 


CTl 


en 


co 


w 


CO 


CO 




CO 


CO 


+j 


+-> 








+j 


^_, 




a 


3 


£ 


£3 


3 


3 


a 


3 


cj 


cu 


(U 




CU 




cu 


cu 








O 


CJ 










CJ 


u 


u 


Sh 


u 


i-l 


Vh 


in 


._ 


cu 


CU 


cu 


CU 


CU 




CU 


cu 


ft 


ft 


ft 


ft 


ft 


ft 


ft 


ft 


c 








O 





O 








GO 


i^ 


VO 


l-O 


^ 


OO 


CN 


M 


<U 


cu 


CU 


cu 


cu 


CU 


cu 


CU 


i-i 




(h 


>-. 


u 




!h 


in 


3 


3 


3 


a 


3 


3 


3 


3 










-4-> 








X 


X 


X 


X 


X 


X 


X 


X 


















§ 


§ 


S 


§ 


§ 


S 


s 


§ 



fa 







M 






u 













£ 






>> 






> 






rf 






cu 


c 





,3 


■<*■ 


Tt- 




CM 


0) 


£ 


CO 


CO 


u 


a 


rt 


T> 


>, 


>> 


Pi 






O 


' cJ 


TJ 


3 


3 




3 


3 


3 


O 


O 


■^ 


HH 


>+H 


OJ 


CU 


CU 


CO 

3 


3 


3 








rt 


rt 


1-1 


CO 1 


CO 


""* 



fa 



fa 



fa 



fa 



pq 



3 


3 






X 


3 


6 


"" H 






„ 


73 


-d 


3 


£ 


a 


aJ 




CO 






bo 


be 


s 


3 


m 








rt 


rt 


fa 


pq 



T) 








































































03 


TJ 


































PQ 


O 




O 


in 


ID 







o\ 




t^ 





lO 













cu 




1-1 


M 


m 


fO 




\o 




4 


t^ 


CO 










PI 


fa 














M 




1-1 


m 


1-1 












> 










































































bo > 

cu o 



Q 



< « 



co 


































■d 


























O 








3 


















pi 


1>- 





ro 


O 


P) 


P) 


f^ 


















i-^ 


PI 


CO 


PO 


PI 


1^ 


0) 


t^ 


fa 




















M 




1-1 


PI 


M 










































































i—i 



































^ a 



1- 

cu 






PI 




LO 


On 




VO 


O 


r^ 




CO 


PO 






1-^ 


in 

pi 


in 


3 


IT) 


l^ 


co 


M 


"* 


VO 


1^ 


t^ 


exj 


VO 


CO 





t- 





On 


00 


O 








M 


M 


1-1 


M 


1-1 










M 


M 






1— 1 



































H ^ H 

>h 13 H tj 

«i W 13 b 

i-l H W m 

fa 



ft 


ft 


ft 




























fi 


B 


e 




































t^ 


PO 


NO 


pq 


NO 


PI 





pq - 


O 


\o 


NO 


VO 


PI 


CJ 


CJ 


CJ 


PO 


■sh 


■<*• 


in 


in 


m 


^f- 


1-1 


P) 


PO 


in 


On 


On 


rh 


00 


PO 


pq 


pi 


pq 


N 


pq 


pq 


PO 


4 


in 


in 


PI 


PO 


ro 


pi 


pi 


PO 




























pq 


pq 


M 





























M 



Shhh 



"A 



P ^ 

H W 

H k W 

Oof^> 



O w 



< 


W 


Ph 


« 


m 


r^j 





^ 


< 


<! 


P 


■Jl 


J^ 





PO 


■<t 


m 


NO 


t^. 







P) 


tx 


00 


On 


O 


VO 






PO 


PO 


PO 


<cf- 


-t- 


^J- 


in 


10 


in 


VO 


VO 


pq 


pq 


pi 


pi 


PI 


Pl 


N 


Pl 


M 


PI 


PI 


Pl 


PI 

1 



THE FOUNDRY SANDS OF MINNESOTA 



I 

c/3 
H 
W 
t> 

a 

i— i 

PQ 
A 

<: 

C/3 
O 

i— i 

Q 

o 

o 
W 

H 
O 
5? 
W 

H 

C/3 

W 
t-5 
i—i 

W 
H 



> 

W 
i-l 

pq 

H 



U 





















■+ 






















on 






G 
o 

In 


^ 














01 




o 


o 


73 

3 


03 
S-, 
G 


3 


03 
3 


03 
U 

G 




T3 
G 




>> 

<u 


CD 

u 

bo 


g 

o 


03 


03 


03 


o 


o 




03 

oo 
On 
01 

CD 
!h 

3 




bo 

+-> 
03 


+-> 

03 
bo 


u 


bo 

G 

3 


bo 
G 
'.5 


bo 
.S 

15 


bo 
_G 

IS 


bo 

_G 

o2 




CD 

N 

G 


bo 




bo 


03 


03 


03 


03 


oi 






- 1 


u 


Vn 


02 


03 


02 


02 


02 




X 


02 




o 


o 


!h 




Sh 




Vh 




B 




o 


M-i 


<+-l 


1) 


CU 


<U 


CD 


<d 




















-+-» 














<-H 




Mm 






V 




>-H 




X) 


T3 


OJ 


03 


03 


03 


03 


X) 




T3 


G 














ft 


G 




G 


03 


rt 


00 


00 


00 


■* 


lO 


a 

03 


o3 


G 




1/3 


03 


o 


O 


o 


tx 


i^ 












01 


01 


oi 


01 


<M 








bo 


bo 












N 

g 






a 




in 

C/3 


en 
03 


t/3 
c/3 


s 


a 


-a 


o 

o 


O 


o 


o 


tu 


dj 


CD 


o3 


o3 


CD 


O 


oS 


rt 


o 


O 


o 


O 


O 


!h 


[H 


[U 


h 


i-l 


1-1 


1-1 


i-l 


1-1 


S 


PQ 


p* 



ffi 



C/3 



T3 
CD 

0* 
03 

pq 

G 

CD 

> 

o 


03 

G 

G 
o 
Pm 

1— 1 


O 
OO 




o 
o 


NO 






t^. 




o 

VO 


q 
6 


q 

6 




CD 

S 

< 


03 

G 

G 
o 
Ph 

G 

i— i 




o 
4 


tx 


OO 






o 

CO 


q 


04 


OO 

vd 






G 
<D 
CD 
Jh 

o 
1 


<D 
U 

G 
G 
O 
G 

h- 1 


no 


tx 


oo 


co 

CM 


q 


o 

NO 


o 

0} 


q 

0) 


o 

oo 


00 

4 

0} 


q 

0) 


q 



>. ^ H M 

Ph 



o 


NO 


rt- 


0) 


o 


o 


^3- 


\o 




OO 


r^ 


VO 


IX 


o 


01 


00 


04 


oo 


to 


Tf 


LO 




M 


OA 


m 


**■ 







H H H 


U 


,^K 


H 


wwo 


en 




O 


O 05^. 



Ph 



01 

On 


On 


00 
00 


o 
o 


o 
o 


o 

o 


o 
oo 


o 
co 


o 
o 



O 

H 


o 


W 
i-l 


<J 


M 




« 


oa 


fei 


O 




< 


< 


f 


C/J 


M 





oo 


On 


o 


00 


^fr 


io 


NO 


ix 


o 


VO 


NO 


(^ 


00 


co 


oo 


00 


CO 


On 


tN 


0) 


M 


oq 


M 


NN 


0( 


0) 


M 



PERMEABILITY TESTS 



89 



Q 
O 

U 






73 t>> 



V- £ 



-d U 

6 ? 
S .is 



<5 



73 




el 




rt 










<!) 


i-< 


S 


O 


nt 


U 


CC 



to 00 _<u 



t3 m . 

u <u ,-T 

Rt +■> S 

J3 „ _j3 

t3 ^ 'ii? 

CD bo Ji 

S 73 73 

n! cd (ij 

u G £ 

-v s a 

a> rt rt 

s >< x" 

-d 

<u M 53 

in rt eg 

cd 

a «J v 

S § B 

<U eg 03 

-m to co 

X) tT tT 

rt cj rt 

CO CO CO 



J3 _o _o g 
ESS I l-l 



73 
Sh 73 



8-3 



73 73 
3 w 



S M 



pq 



73 73 

CO CO 



h h 



73 £ 
<U 73~ 

82 

>— ' o 

.2 S 

O ^ 
„ o\ 

5 co 
2 rt 



S3 S3 

73" 13 

el G 

rt rt 

co co 



fn h 



u 



>)>>>,>. 



s s s 



t)- ^O 00 o 



0000 



CN * (O CO 

73" 73 73" 73 

CD OJ CD CD 

N N N N 

'co CO 'co 'co 

cd cd*" cd~ cd~ 

"n n 'n n 

f!l Cj ci CS 

S3 S3 S3 S3 

C cr 1 ty o* 

73 73 73 73 

CD CD CD CD 

.d ^3 ^3 rG 



CO CO CO CO 

bo bo bo bo 

a a el el 



CD <U CD CD 



>> 73 

S-i CD 

<D +-\ 

to bo 

§ "H ■ S 

r_ ' rt co 

J "8 s 



^ J; 



U2^ 
« 1 •- 

S3 S3 rt 

1 < pq 

o 

rh, CO CO 

^ rt rt 

C CD CD 

oEG 

Vh rt ai 

<-H CO CO 
CO * (B 

co co co 
CD CD CD 

000 






a ° ° g 



co h w Q 

co O Ph £- 

O S 

S3 -I 

o t* 

« (J 
W 

H 






t^ 


t^ 


CN 




ro 



co tj- co 10 



H « N Q 

m W g g O 

ri p w ^ » 

m ^ u v! < 

< a « ^ c/3 

W ^ W pc, c 

« ^ M -A 



Ph 



M O 



Clh 



^ -t 




ro C\ 




n 


l-T) 


vo 


C^ 


t^ 


\C 


M IO 


U-i M 


co c^i 


3 -t- 


VO 


O 00 


CN 


CM 


CO 


^o 


CM 


LO 


W 


06 ^^ 


CO 1^ 


m ro 














CN 




ro 




1^ vo 


w 





s 



fe: W H 

5 3 w 

^ H « 

r^ * S 

H U > 

a « ^ 

O W >* 

U Ph M 



" I 



lo o o o o 00 



£ 



U 



^ 



w C| CO 
K lO VO 
CO CO CO 








C3^ 


t-l 


CO 


^ 



M 


CO 


10 


t^ 


VO CO 











10 \o 


CO. 


CO 


CO 


CO 


CO CO 



<! 


w 


QJ 


pq 







< 


t> 


J ^ 



, 73 

^ el 

el rt 

rt co 



< PQ 



X >i N 



pq u 



< pq 

0\ !3> 



<; pq O Q <pqu 



90 



THE FOUNDRY SANDS OF MINNESOTA 



I 

(Sj 

H 

W 
H 

I— I 

V 4 



> 
w 

pq 
< 















u 


iH 




Td 




C 


























o 


OJ 




<L> 




(U 


























<+4 






CO 

<U 




tu 

5h 

bo 


























_p 


3 




"^ 






























(0 

u 


o 




2 

bo 




Td 
<u 

to 












R 
tu 














£ 


<u 




~ 




(U 












tu 














u 


p 




Td 
















bO 














^ 






<u 






























p 
o 


.p 








o 












Td 
CU 




H 










8 




g 




o 












to 




CO 










<+H 






rt 
















CU 




H 










^2 


p 








Td 
P 












id" 




Ph 
O 










CO 






Vh 


be 


to 

M-i 


bo 






nd 




rt bo 

-^ .8 




CO 

o 






>> 


>> 


o 
o 


•S 




p 
o 

M-l 


.8 
"co 


o 

<LI 


_P 
to 






cu" 


P 


*-" QJ 










1h 


(h 




>^ 






tu 


o 


<u 






o 
o 


<u 




H 

p 

o 
O 

Q 
< 








T3 
P 


_R 






_P 


tu 


p 


R <u 


u 

td 

tu 

CO 

o 

O 


p 


u 

bo 


tU cu 

u 








p 
n 8 

O rt 

M-H tO 


P 
O 

8 

to 


8 


P 
P 
o 

8 

to 


I) 

to 
O 

o 
>^ 


id 

OJ 
5-i 
<L> 
P. 

g 

tu 


u 
o 

M-i 
CU 

tu 


o 
to 

pq^ 

Td 


1 S5 
T3 to <u 


CU 
CU 

1-1 
bo 

id 
tu 

CO 
CU 


00 

oi 

01 

Td 
p 


Td 

CU 

to 

<u 


Td o 

<U Mh 

p <u 

^ V 

u <u 

pq ^ 




R 






>> >> 


X 








.. 






w ^ 


- 

CN| 
OJ 
01 


■^ 


01 


2 




< 






^3 -P 


.a 


rt 




>^ 


^ 


< 


°-0 n 


Td" 


01 
01 


*o 


TJ 3 




C/3 






(LI <U 


•73 


to 


<U 


02 


'rt 


m 


2 <u "P 


(U 

nd 


to 


8 


P "0 

03 —< 




O 

w 

H 






CO w 

"S 'S 

5-i 5h 
P P 
MH M-l 


CO 

'p 

P 


u 

T3 
P 
P 

o 


8 
<u 

o 


P 

o 

M-l 

"a; 
tu 


< 


u 

Td 

P 


-^12 P 

" < pq" 

R 


u 

p 


"o 

8 
< 


^. 

Sh 

•"O 

R 

P 


< 

01 
01 


< U 

01 01 
01 01 




< 










MH 


o 


CO 


< 


to 


o 


O to to 


o 


to 


o 


to 


CO CO 




15 






T3 TJ 
1) <u 


id 






8 


rt 


Mh 


u a rt 


M-i 


rt 


MH 


a 


o3 o3 










tj T3 


2 


<u 


< 


8 

o 


OJ 


<U 


<u 8 8 


tu 


CU 


V 


<u 


cu cu 










'o "o 


'o 


8 


o 


8 


8 


8 


8 


8 


8 


6 8 










6 6 


8 


aj 


to 


>+H 


5-. 


rt 


rt 


»h n! rt 


rt 


cd 


rt 


03 


0) OJ 










to 


a! 




M-i 


CO 


to 


bo to co 


CO 


CO 


CO 


to 


to to 










► 


„ 




!^ 


T3 


„ 


























T3 T3 


T3 


"d" 


CD T3 


R 


'Td 


R T3 


nd 


T3 T3 Td 


Td 


Td 




03 


d o3 










P P 


R 


p 


T3 R 


rt 


P 


tu p 


R 


RPR 


R 


p 










rt rt 


rt 


rt 


p rt 




rt 


<u rt 


03 




rt 


rt 


5h 


!h 


!h 5h 










co to 


co 


CO 


■ w to 




to 


Sh CO 


to 


CO co to 


to 


CO 




















^5 


N 




bJO 
























to to 

co to 


CO 
CO 


u 


tt> 

8 


3 


to 

CO 


to 

to 


to 

to 


»H i-i Vh 

o c o 
o o o 


O 
O 


5h 
C 

o 


8" 


8 

03 


8 8 

o3 o3 










Vh »h 


1h 


o 


rt 


p 


Jh 


5h 


!h 




o 


o 


O O 










pq m 


pq 


U 


C/! 


a 


pq 


pq 


pq 


SEE 


E 


E 


J 


1-1 


J J 




« 


H 
co 
































IX! £ 


£< 


H 
































co Ph 




p 


Ki \l" 


o 


o 


VO 


LO 


■* 


oo 


lO 


O " ^t- 


tx 


tt 


•+ 


00 


o o 


g Oi 


co o 


W 


£ 


■3- VO 


^o 


oo 


U"; 


tN 


oo 


LO 


VO 


U-y OO 00 


H 


o 


-1- 


01 w 


M W 


< ™ 


o 




h \o 


oo 








CO 


OO 


t^ 


i-i oo oo . 




w 


CM 


NO 


CD l-^ 




° H 


o 


3 




























of 


y to 

2 ° 




U 






























^ § 


J 


H 
































!* 


, 




>< 






























M W 


W r 


P 


H 


t-^ t^ 


c^ 


M 


VO 




u-> 


o 


01 


vO O "^ 


in 


UO 


01 


oo 


oo 01 


rf Q 


(5 


i-l 


t>. ro 


<N 


M 


4 


O 


VO 


!>. 


ro 


>0 oo NO 


■* 


OO 


CM 


OO 


M OO 


M ^ 


s £ 


«J 


M 


i-( 




ro 


Tl- 


O 








HH 


M 




hH 








A & 


OT 


< 










i- 1 




















Ph O 


































H Eh 




< 
































Ph O 


< 




Ph 
































S 8 

& S3 


H 
































W 
M 
P 


o 

M 












O 


? 


* 


O 


* 

O ^O oo 


•o 
6 


O 
6 


o 


O 


o o 


H 


w 
































CO 


H U 


£ 
































O 


£ « 


































s 


o w 
U Ph 




































CO 

OT R 

H 5 


P 


CO 


tJ- 








00 


VO 




•"t 


o ^t- 




\o 


o 


ti- 


rt- w 


<c< 


< 


t-^ 








On 


o 




On 


O NO 




On 


uo 


ro 


oo o 


o 

hH 

H 


c/3 




ro 








OO 


■* 




oo 


Tt oo 




oo 


OO 


OO 


OO OO 


£ 


Ph 

o 


& 
































* Ph 

g ° 

H « 


CM 




<5 PP 


u 


< 


M 




< 


pq 


U 


< pq u 


< pq 


< pq 


U P 




° 3 


< 
x 




^ f^ 


t^ 


oo 


00 


o 


h-t 


w 


M 


0) tM 01 


OO 


OO 


^t- 


tl- 


rh •* 






M M 








M 


<N 


CM 


CI 


M « tN! 


tN 


01 


CM 


Ol 


tM tM 








tM f"l 


M 


H 


CM 


<N 


tN 


CM 


N 


0} tM m 


01 


Ol 


CM 


<M 


01 01 




J £ 



































PERMEABILITY TESTS 



9i 



S fl c c 

cl> <L> oj a; . 

(J U t) J) ^ 

bo bo bo bo 'd 

T3 13 -d 13 T 5 

<L) <D CD U U 



T3 T) 13 13 "O 

oj u «j u m 

T3 13 13 13 13 

"o o o o o 

a a a a a 



Tf Tj- -t- -t- -r 

M Cl Cl Cl Cl 

N N « fi N 

saess 

rt k! « rt ™ 

o o o O o 

I4_l l+j «+H <-H >-H 

o o o o o 

o o o o o 

1-1 eo -*)r in 

T3 13 13 13 13 

s c S c a 

rt cd 03 rt ™ 

(O M fO M tO 

Cl 0) 0) CI CI 

CI C) Cl Cl 

■-d 13 13 13 13 

c a s q q 

03 03 rt 03 rt 

t/a CO U3 02 03 



00000 
qz; cm cfi cm c£ 

Lt_, M-( *-H »-M M-( 

00000 

^^^^^ 

o o o o o 

O 00 l^ VO 10 

<d <lT <d iT <d 

J_ )_, in U U 

3 3 3 3 3 

^S X rS rS rS 

§ S S § ■ § 





3 


c 




<u 


OJ 




u 


u 




S-l 


Sh 


_, 


bo 


bo 










13 


13 
<L> 




Cfi 




13 






1) 










13 




13 
<u 


13 




13 


O 


13 





a 


13 


a 


£ 








a 


>> 


13 


„ 


13 


3 


•+ 

Cl 


S3 
3 


<-H 


Cl 


O 


<u 


„ 


4-H 


a 


CO 


<L> 


TO 


CI 


a 






d 


VO 


CI 
Cl 


en 


Cl 


Cl 


m 


13 








to 

a! 


Cl 


to 


>, 


13 


IT) 


u 
13 


to" 


Cl 


3 






3 
O 

4-1 


Cl 


-1- 

Cl 
Cl 


<u 




r-j 


a 

to 


.3 


'% 


en 


5 


13 


13 


13 


CD 


OJ 


<u 


3 


en 


en 




3 


3 


£ 


j£ 


E> 


13 


ni 


TO 


03 


Vh 


in 


en 


1 - 


^ 




1 a 


£ 


a 


c3 


TO 


o3 





O 


O 


1 »-i 


J 


l-l 



Ph PL, Ph PL, PL,' Ph 



C/3 



N(nNcnNt/3Nc/3NcoN 



eg a 



N> +H Ni+h N<-H N 4-1 



3 

a 



3 
a 



a 



a 



3 
a 



13 O 

■— ■ 13 

en S3 



- S3 

23 rW 



£ <J 



pq 





en 

TO 




en 
to 




en 

TO 


01 




Tl 




+-> 







»-. 




^H 




"ni 


s 


to 


2 




a 


a 


ol 




to 




01 


c/j 


13 


en 


13 


to 




^ 


<L) 




(LI 




13 


13 


H 


13 


H 


"3 


S3 


ol 


a 


3 


a 


S3 
rrt 


to 


</j 


oJ 


en 


rt 


w 



ft, pu pq pL, 



M W 



w « 

5 w 2 © 



w 

g en o 

M O H 



w 



J. 



w 



O M 

o ^ 

M 



<M 00 O CO 1-1 

en a 01 ^ h 

Tf Tf- CO CO CO 



-h 


t^ 


in 




r^ 




co 


Cl 


Cl 



H Pi 



Z Q 

1-1 o rt < z 

H h K l ^ 

«S j M -^ C/2 

W G W rx, _ 

§U Ph o g 

§ h * « < 

o - ^ 



Ph 



Ph 



w O Cl 


H O 








00 





m 


VO 


H 





00 


10 mvo 


1^ 00 


00 


0\ 


C\ 






CO 

1^ 


CO 


CO 


CO 


l^ 



ci in 

M \0 



w 



t5 w H 

5 o w 



•3 H H 
H ^ S 
en W W 



^ 



o w >^ 
U Pm m 



w 


en 


n 


£ 


rl 


W 


<! 


H 

w 


H 


C/3 



u 





o Z 



0\ 00 ■* 
CO co CO 



t^ 


CO 


00 


f^ 


00 


CO 


CO 


CO 


■* 



O ° 




<KUfiW 






1-1 


■xt- -T -H- Tf •*• 




a pq 




d ci ci ci c. 


in 


§ 


§ 


Cl C'l ci cq Cl 


Cl 




< 


CO CO CO CO CO 


Cl 


C/2 


Cl Cl Cl Cl Cl 
Cl Cl Cl Cl cq 





<o 


f^ 


M 


Cl 


M 


Cl 



<! 



pq 



U 



fl 



w 



h 



<J pp 



u 



9 2 



THE FOUNDRY SANDS OF MINNESOTA 









bo 


Tl 






bn 








S3 


u 






S3 








































U3 . 
















19 


u 






49 








+-> 


ft 








Tl 






1) 


>s 






09 

Sh 


49 






o 








O 








<+H 


T3 








73 

(U 


c 


S3 


4J 

42 


if 


S3 


a 


49 

42 


ft 




<u 




19 








Sh 




X) 


o 






T3 


49 


t* 


bo 


<L> 


o 


bo 


be 


19 


71 






Sh 










O 




Tl 


Tl 


r, 


Tl 


rtrt 


~ 


D 












a; 




' ' 


09 
19 


71 


Sh 


'ft 


i/i 


03 


U 










qj 




>-. 














ci 


T3 


T3 


^J" 


Tl 


S3 


Tl 


+J- 


Tl 






— 


U 






4j 




49 




be 


rt 


09 




bo 


rt 


4-1 


Uh 


43 


it! 


43 


43 


4= 


ft 


T3 


T) 


X) 


T3 


Tl 


t3 


Tl 




S 


19 


<u 




09 


<u 


19 


s3 


S 


B 


H 


B 


s 


s 


^ 


id 


3 


S 


-T 


B 


s 


B 






n) 


03 






rt 


CTt 


its 




u 


Sh 


-t) 


sh 


Sh 


)h 


T3 


T3* 


T? 


np 


^ 


Tl 


Tl 


rtrt 






c; 


1) 






a; 


q; 


o 


"d 


-73 


t3 


o 


T3 


T3 


rtrt 






O 


o 


rt 


a 


O 




jJ 


s 


B 


09 


B 


s 


B 


crt 


~ 




„ 




„ 




>i 


< < <u 


f*. 


< 


<; 


<! 


Sh 

X) 


•* 


-T 


^ 


U 

T3 






0) 




(N 


0< 


^ 


01 


N. 


0) 


y 








3 








o 


03 


r/i 


79 


O 


r/> 


r/i 


71 


MH 


rt 


rt 


rt 


<-H 


rt 


rt 


rt 


<u 


<U 


49 


09 


u 


(19 


(!) 


a; 


b 


B 


b 


B 


B 


B 


£ 


B 


r^ 


rt 


rt 


rt 


rt 


c^, 


rt 


rt 


</; 




03 


09 


09 


09 


09 


03 


T3 


T3 


— 1 


-d 


T3 


-rt 


T) 


Tf 


53 




S3 


£ 


rt 


S3 


S3 


c 


rt 


rt 


rt 


rt 


rt 


rt 


a 


^ 


73 


73 


03 


03 


09 


03 


73 


73 


Sh 


Sh 


)_, 


Sh 


Ih 


u 


in 


Sh 


o 


O 


C 


o 


o 


o 


O 


O 


o 


O 


o 


o 


o 


rt 


O 


rt 


















fa 


4h 


fa 


fa 


fa 


fa 


fa 


4u 



T3 "- 1 

09 D 

•4-J t3 

OT s 

19 2 

"^ 42 



T3 

* S 

19 rt 

T3 
S3 " 






19 42 

T3 rt 

rt """ 

S S3 



U 



bO "l 

.5 °° 
"w _ 

LO Sh 

0) T3 
S3 

■§3 



s 


ft 


o 




Mh 


S3 






09 




S3 


Sh 


rt 




rn 


-rt 










.s 


42 


19 


r/1 


T3 


rt 



^ 5 



- T3 

■* rt _, 

01 79 ^ 



O 



r/i 


rt 


rt 






Tl 


>> 


rl 






TJ 


rn 


S3 




P 


Sh 



q3 5* 



bo r ; 

S3 ^ 



>, S3 S3 6 

Sh 19 aj t(_i 
fl D 09 (J 



T3 ? 
09 U 





19 


19 


19 


—- 


-rt 


Tl 


r. 


a 


rt 

£ 


71 


42' 




T) 


03 


u 


S3 


<A 



19 Sh 

T3 .2 

S3 S3 ^ c 

rt G S3 



U 



M 



U 19* 
> 

2 '35 
3 



19 

fa/3 

.5 "£ 

^ OJ 

>, B 
5 I 

^ (19 

T3 ^h 
S3 O 

s B 



79 " 

19 ^ _ 

+J T3 

Sh & 

+r -S 4«s 

-H rt rt 

be ^ ^ 

T3 j 

19 ., *" 

s -o - 

« Sh T3 

S3 rt Sh 

rt ^ 2 

Sh 43 

_r t3 T3 

T3 qj 19 

T3 S B 

•3 B B 

C rt rt 



fl <J <J pq pq 

O 79 79 79 73 

^ rt rt rt rt 

19 19 19 19 

£ B B S B 

>h rt rt rt rt 

bC 79 79 79 79 



^ 4h rt ^ 



u 



T3 T3 Tj T3 T3 C 

5 S3 S3 S3 S3 • « 

d d d d d 73 

79 03 79 79 03 <JJ 



o o o o o 

o o o o o 

fe fe fe E fa 



B ^ 



<J 






tn 



•H 73 
2 O 

^ o 



J H 



O KH 

o s* 
« (J 



1^ M LO Q\ 



m <o >t tq 
O t^ ro M2 



O U"> OO lo 

io oo oo m 

to M m ro 



s 

< 


w 

n 

S5 

*-* 


Iz; 
w 
u 


•< 
tn 


Q 

CA9 




> 


HI 

Ph 


o 




Hi 

w 

4h 


fn 

o 




w 
o 

< 


<! 








w 


E-i 




w 




C3 


w 




« 
^ 


H 


<! 
H 


o 






K 


5 


w 




r/3 


W 


W 


> 




IH 


H 


U 


[> 




O 


^ 


« 






S 


O 

u 


Ph 


PQ 



M C\ in ro 

M 4- lO VO 



M Tf fl O 



m ro t^ ^r 



i-i 00 ro oo' VO* 



O N K N 



o oo oo oo 



O O O ro O 



5 S « § 

a w c < 

£ Sh tf ^ 



o 



n oo 

IT) 9^x 

OO OO 



in VO vo 
oo oo oo 



tx o m r^ 

^j- VO y}- ^c 
oo oo oo oo 





45 
41 


< pq u q 


< PQ U P 


« pq 


§ 


H M HH hH 


M 01 M 01 


S S 


^- •* TJ" xh 


■*■*'+'* 


^ ^ 


c/a 


0) OJ 01 M 


PI N N « 


J ^ 









<^ 


PQ 


pq PQ 


u u 


u 


u-> 


VO 


f^ 00 


C\ o 


H 


^t 


-t 


^ Tf 


■^- io 


lO 


cvi 


M 


01 cs 


M N 


M 



<fflUQW 

N N N N N 
U~) \J~i ID ITi xn 

H d N IN N 



PERMEABILITY TESTS 



93 













Vh 
X) 






id 




xT 

2 


u 
d 
ft 




XJ 
S3 
03 




XJ 
<U 

to 
CU 




xT 

CU 
X 












03 




1H 


I) 




a 


M 


>, 






+J 






a 












O 
O 




T3 


>) 






00 


X) 

S3 

3 


O 




xi~ 

1) 




















S3 






03" 

13 


'x 




2 




N 












in 




to 


oT 




S3 
u 


'a 





N 

S3 






a 




3 













<N 






to 




s 




_S3 


O 








Vh 
















O 


o 




O 


>, 


Vh 








.O 












w 




O 


© 


^ S3 




^3 




^2 




u 










H 






rt 


fl S3 


q3 




<u <u 






X! 






<u 










to 








w <u 






(U V 


03 


<u 


<L> 


Vh 




^ 




3 






W 






>> 


D cj 


to 


X 


u u 


03 


X) 


03 


O 




3 










H 

O 
to 






Vh 


Vh !h 




1/ 


be be 


03 


rt 


3 

03 


<-H 




03 














XI 

S3 


be be 


ft 
oo 


to 
V 


Td td 


^2 


a 


X) 


OJ 


^ 


3 

CU 


bo 
pi 










3 

o 

<-H 


T3 T3 

^ ^ be 
tn to c 

^ ** « - 

- ^ <u 
T3 •£ -*-. 


o 
o 

to 
+-> 

ft 


iJ Ji be 

to to S3 




S3" 
1) 


rt 


S3 
03 
to 


Vh 

bo 


s 


CU 
Vh 

be 


"0 
03 

M-H 






O 

H 
H 






£ 

d 
03 


"-H 

o 


"^ "^ to 

X -g +J 
i3w « 




a 




1) 

be 

X! 


"ft 

a 

03 


be 

c 




X3 
CU 

CO 

OJ 


.a 
5 


XJ 
CU 

03 

CU 


Vh 

O 

>+H 






s 

O 

u 






bio 
-3 


rt be <u 

r^-l r^< * 

T3 13 D 
<U CU o 


3 


J3 .rt w 
•« -^ o 

x) Td dj 
S 8 <u 

Td 
id x) u 





> 





0^ 


03 

03 

03 


xT 

CU 

XJ 


bo 
3 


J2 


be 


X) 
CU 
03 

3 






Q 

< 

Q 
15 






3 
3 
o 

4-1 


26^ 

S3 S3 (U 


oo" 

lO 

CI 

Td 

S3 
c3 

lO 


H 


cj 

OJ 

bo 
p S3 


X 

a 


X3 
S3 
d 
03 

XI 


XI 
<U 
03 

cf 


"o 

a 

bio 


cf 

c^ 

CS 
03 

d 


X) 
CU 

a 

a 
03 


CU 

CJ 

3 
03 

Vh 






•< 






.£f 


•OTl tl 




x) xi rt 
*o IS ■£ 


03 


. 


CL> 


Ov 


>^ 






C/3 

O 

H 






Vh 

o 


'o 'o ■£ 

a si 


£ 

c 




03" 
ft 


u 

O 


U 

d 


<M 

03 
03 


03 
03 
CJ 

3 


Vh 
X3 
3 

3 



Vh 

3 

^2 


a 


Vh 






H 
< 






o 

Vh 

1>> 

CU 

u 

be 


< < PQ 

tO tfl 03 

o3 rt rt 
<u <u <u 


Td 


a 
ft 

M 

t5 


tO 05 03 

cd cd rt 
<U l) CJ 

$3 a a 


O 

03 

X) 
2 S 


a 
a 

S3 


ft 

CO 

OV 




Vh 
X3 

S3 
3 
O 

>+H 

CO 




Vh 

*d 

H-> 

3 


a 

03 

03 

XJ 


< 

to 

03 

CU 


X3 
CU 

Vh 
O 

3 v 

Vh qj 
Vh 

^ be 

XJ 












£ £ S 

rt oj aj 
to to to 


to 
Td" 


to 
N 


d d d 

tO t/3 03 

T) T) T) 


^-1 qj 

h bo 


O 

xT 


X) 
S3 
03 
03 


a 
0! 

03 


a 

oJ 

3 

Vh 


to 

3 


a 

o3 

03 












id 


12 tT Td xT 


S3 




S3 S3 S3 


xT Td 


S3 




xT 


O 


T3 3 


X~ 


3 XI 














ii S3 S3 S3 
■*-» rt ctj o3 


to 


1h 


as rt rt 

03 03 03 


S3 ^ 

CS I/) 


to 


O 
O 


3 


Vh 


3 <u 
03 <U 


3 
o3 


03 CU 
1/1 to 












to 


to to to to 






O) (J 




to 


O 


03 Vh 


03 


H^ CU 














<L> 


bo 


o" 


bo be be 




N 
S3 


53 




M-l 


bo 




cu 43 












o 


P< )-, Vh ^, 

o o o 


_S3 




.s .a .a 


03 

03 




Vh 

O 




to 
03 


to 

03 


3 












o 


o o o 


'o 




'o 'u 'o 


d 


O 




O 




o3 


cd 


O 














cS 




rt c3 rt 


u 


Vh 








Vh 


Vh 


Vh 










K 


E 


E E E 


h 




h hh 


pq 


pq 




E 




pq 


pq 


pq 






OS 


H 
to 


































< 


H 


Ph 

w 




























to 


fe 




Q 


cq 


CO w M 


oo 




W K "t 


00 


in 









t^ 


00 


CM 


1? °* 


to 


o 


M 


|z; 


ro 


\f M O 


MD 




W t^ lO 


i^ 


o\ 




00 




f-i 


l-H 





< 




o 




ro 


■<*■ CO ■* 


t-N 




lO M ^t" 


CO 


CO 




00 




CO 


CI 




- P-< 

M 5 


O 

H 


co 

fi 

w 

H 


3 
o 

K 


2 
1* 

u 


M 












M 














H § 




iJ 


H 






























>H 




, . 




>H 




























H « 




n 




H 




























M W 


g 


p 




o\ 


*o 00 fl 


CO 




ro w in 


00 


CO 




CO 




0> 


■* 


in 


Hi O 




£ 


3 


H 


m n^o 


CO 




■^ rj- m 


ci 


M ' 




cj 




l^ 


H 


4- 


u 


<! 


P3 










M 














1-H 


CM 




tA 


C/D 


W 


< 




























W 

Ph 


o 


































fc 




< 


Ph 
w 




























Pm o 




<! 




Ph 




























H 


M 


W 
O 
<! 

15 








^ °i * 






"P "7 * 




q 




vo 

O 




XI 

CU 


X3 

cu 


XJ 

CU 


13 
H 




o 




o 


Ch On CO 


o 




oo oo vo 





00' 




co' 




a 


3 

a 


3 

a 


03 






£ 
























^"* CU 


^~* CU 


*^ CU 


O 


« 


M 
































^d 


O 

U 


H 


>< 
























CU 


CU 


CU 


r^ 


Pn 


pq 
























XJ 


XI 


X) 


W 

o 
w 


H 
w 
W 

O 


en 




in 


VO 


tM I s ! 


"+ 




In. O f^ 





00 




C) 




CO 


ci 


CO 


O 


< 

o 


<i 


Tt- 


-+ CO 


CO 




in M ^t 


00 


vo 




0\ 












2; 

M 
1-1 

u 


PS 

o 


CO 


CO ro 


CO 




CO CO CO 


co 


CO 




CO 




Tt- 


■* 


CO 




> 
Pi 
O 
H 

<! 


O 

Pi 


i-i 

Ph 




< 


pn u P 


< 




pq U P 












< 


pq 






Pi 


pq 


§ 




00 


00 00 00 


C\ 




Oi Oi Oi 


O 


C) 




CO 




vo 


VO 


1^ 




O 


S 




LO 


lO ir; IT) 


lO 




in >r, m 


VO 


o^ 




o\ 




C\ 


Ot 


C\ 




< 


13 






es 


CI CI CI 


"1 




cm ci n 


CI 


M 




CI 




CI 


CI 


N 




1-1 


fc 

































94 



THE FOUNDRY SANDS OF MINNESOTA 



< 

co 



+-> 


pi 


o 


en 


a; 


3 


+J 


<U 


> 


o 








rrt 


rl 




rt 


U 


^3 


3 




TJ 






rt 

Xi 


60 


Pi 


TJ 


o 


Pi 


rt 


Pi 


CO 






n 


tn 


O 
io 


>> -a 




O 




u 


<L> 






j-i 


■n 




nt 




OJ 


p 

p 


y 


£ 


u 

Tl 


.a 

a 




- 


P! 


p 
3 


3 


rt 


in 





& 



O xJ 



W 



u fl .2 



^ S 4 rt 



o 



*£ >> ni 



£ "5 



PI TJ 
•- P 



o tJ 

■^ PI 



TJ M-l 

PI ° 



o <u 



rC ° Si 



o 



lO 


00 


<+-i 


<N 


o 

m 


o 


o 




o 




Tl 






P 
rt 


rt 


■d 


Vh 


pi 






P 
o 


o 


,G 


ft 


LO 




S 




P 




a; 


,Q 




!-l 

o 




1) 


U 


VO 
O 


-1 




m 


Pi 







^ «J 



•2 > ^ 

CJ !> TJ 

CD .G 

2 § ^ 

FJ *- i> 

" a J3 



o tJ 





1) 

a 

rt 


M ^ 



O CO 



<u 



1) 

^ tj ^ 
2 pi 

a rt m 
« tJ 

c/0 l,-, rt 

g o a 

VO tn *"• 2 
O rt £ -S 
""> ft £ ^> 

1) 
in 

o 
U 



bo o\ 

P o 



J-, TJ 

ft p 

<u rt 

tJ ^ 

p *° 



O rt ^! 

+j O. (h 

P^ O 



ft 
CD -Pi 

p "^ 



£ a 



2 p £ 
* p 

3 2.2 

2 2 § 



TJ tu 

.a £ 

rt w 

CO ^ 

rt 

TJ 

rd p 

o rt 

« 2 

CD 

P 



-P « 
SP rt 
P 



13 « 



UJ s 

a "3 
j a 



tJ tJ 
P <u 
rt 52 



<u d 

dn .a 

^2 



o M cS 

^ Pi cl 



^3 "In 

d rt 

• 5 a 



>> 2 

p -^ 

rQ 



M P 



<U <-H Xi 



p< 



-a -5 



^ a 



<u 



% U 

Pi 



u 



-d o 

60 n 



S P 



FP 



^ a. 

P rt 



Pi 


-d 


rj 


~ 


Tl 




o 


a 


^ 


o 


P 

o 

d 
^4 


OJ 


3 


^ 


P 


m 


a 
< 

V 

-p 


rt 


IO 


S3 


d 


U 


M 
IT) 


o 
o 

d 


TJ 
P 
rt 
tn 


o 

p 












rn 


lh 


t^ 


r/i 




<U 




o 


*-* 


<U 


Tl 


J3 


a 


>! 




O 




H 


u 


-d 
P 

rt 


-d 


d 


d 


a 

rt 
en 


d 






- 


i^ 


o 


vo" 


a 


lH 


IM 


~ 




M 






Tl 


rt 


w 


IO 




f-1 


rl 


CD 


^J 


^ 




a 


rt 


d 


^p 












IO 


-d 

53 


o 
o 


0\ 


<U 
TJ 

P 


rt 


tu 


„ 




vn 



Ch '-H VO 4D 



U 



tn rt 



p -p 



TJ -3 

p 



>> rt 



a k^ 



o U 



.2. * 3 



u 



M O 



o 



rt U 
O 



M CO 



ft w 



rt P 
O 3 



' H erf f -1 

a g S w S 



CO 



w W 



r« P4 



Q H h 



P> i-l 
O lx 
« CJ 

M 

H 



oo oo oo 

\o VO "I 



*-" O 00 TtOOO rovO u-> 
MOOMni-iTt-iOlO 



^2 „ 



TJ <U P 



rt .p 



« H Q 

W Z 2 Q 

« « CO co 



w h 



s 6 ^ ° S § 



Ah <=> 



w < rt 



oo 


CO 


M 


VO 


CO 00 


0^ 


t^ 


■* 


0^ 


o 


o 


M 


O CO 


VO 


M 


l-l 




<N 


(N| 


^1- 


<N <N1 




M 


<N 



o> oo oo 

^f- in in 



10 uo t-^ 
m lo r^. 



w C\ t-N C^ LO O 



2: w 

£ O 

W W 

H CJ 

Z * 

o w 

U rt 



O O O O 



OOOOOOOOO 






w 



fn 



o 



§ 


LO 


l-H M 


w 


<J 


oo 


co oo 


-1- 


« 


CO 


CO co 


co 



m CO CO 

VO LO LO 

CO CO CO 



oo oo oo 

CO co co 



t^i-iTj-ncqco^Oco 
M oo oo oo oo oo \D r^vo 
cococococococococo 



o 




H 


« 


< 


w 


K 


M 


O 

n 

< 


P> 


rt & 



On M CO rj" lO 
O O O O O O 

LO LO LO LO IO IO 



LO LO LO 



C^ CO -^ 

LO LO LO 



Lo v ot^ooo\onc^co 

MMHHHlNMCqC^ 

lololololoiolOlolo 



RESULTS OF PERMEABILITY TESTS 95 

RESULTS OF PERMEABILITY TESTS 

Permeability tests are recorded in Table XVI, and include tests of 
clean quartz sand of mixed sizes as obtained from the pits ; the same 
quartz sand sized by the screen sieves to 20, 40, 60, 80, 100, and 200 
mesh ; loams in the dry state and in the molded state ; brass sands, dry 
and molded ; core sands, dry and with artificial bond ; and various mixtures 
of sands, loams, and other materials employed in the foundries. 

The permeability of the material is expressed in Table XVI in two 
ways (a) in the actual time in seconds required to pass 74 liters of air 
through the column of sand 2 13/32 inches in diameter and 3 inches long, 
at constant pressure (see page 43), and (b) in percentage permeability 
in which a standard sand is taken as 100 per cent permeable. 

A few experimental tests were run on the permeability of carefully 
sized sands to test the apparatus and find whether the results agreed 
with those obtained by other workers with similar material, and to find 
to what extent the permeability test could be relied on as reflecting the 
structure in molding sands in general. 

Permeability of sized sands. — Samples 228 A, B, C, D, E, and F 
were sized sands obtained by screening sand from the St. Peter and 
Jordan formations to the sizes shown in the table. The coarse sizes, 
A, B, and C, came principally from the Jordan formation, and the finer 
sizes, D, E, and F, came principally from the St. Peter formation. The 
nature of the material and the character of the grains were essentially 
the same in each case, that is, the sand was essentially pure quartz, the 
larger grains of which were exceptionally well rounded, and the smaller 
sizes less so. 

It will be observed that the 20- to 30-mesh sand, 228 A, was too 
per cent permeable, agreeing with standard sand 215, and that the other 
samples, B, C, D, E, and F, showed a gradual decrease in the permeability 
consistent with the increase in fineness. 

Samples 200 A, B, C, and D represent a second series of sized sands 
obtained by sizing quartzite screenings, resulting from crushing Sioux 
quartzite. The results of this tests were not entirely consistent, since 
the 20- to 40-mesh sand showed less permeability than the 40- to 60-mesh 
sand. Similar discordant results were observed by King, 16 in testing the 
permeability of crushed glass, sized to 20, 40, 60, and 80 mesh. 

The series of samples 223-224, A, B, C, D, and E, represent an 
experiment with a series of mixtures of two loams, differing noticeably 
in permeability, to see in what degree the permeability of the different 
mixtures would reflect the proportion of the two loams present. 

These loams were given 10 per cent of water after mixing and were 
rammed into molds in the usual way. They were tested green, except 

10 Op. cit. 



96 THE FOUNDRY SANDS OF MINNESOTA 

in the case of E, which was tested loose dry. This series is of interest 
in that the mixture of 60 per cent of 223 and 40 per cent of 224 showed 
the maximum permeability of green sands, which was noticeably greater 
than the permeability of either of the constituents of the mixture. 

Many samples were tested both as loose dry sand, pestled to break 
them up into individual grains, and also as damp molded sand. The 
molded sand was in some cases tested both as green, damp sand and 
as air-dried after molding. In general the molded sand, even when 
rammed harder than ordinarily, was more permeable than the loose dry 
sand. Most of these sands were loams in the proper sense of the term 
since they carried a considerable clay content, but they were sands in 
the sense foundries use that term. With some materials, however, differ- 
ent results were obtained, that is, the loose dry sand was more permeable 
than the damp molded sand, as for example in sample 9. 

The difference in permeability of the sands in the loose dry condition 
and in the damp molded condition was plainly due to granulation. Some 
sands, when moistened with water, allowed to stand, and then riddled, 
granulated readily, developing clusters of grains of considerable size, and 
the granules were sufficiently coherent to stand up against ramming in 
the mold ; so that after molding intergranular spaces remained large enough 
to make the mass more than ordinarily permeable. Other sands treated in 
the same manner did not granulate as in the case of sand 9 ; the subsequent 
molding served only to make them more compact than when dry, and the 
permeability was correspondingly less. 

From the mechanical analyses, Table X, it will be observed that sand 
9 is moderately coarse. The percentage of fines, including the silts, is 
small, and the clay content, 2 per cent, is sufficient for bonding purposes. 
Sands of that physical constitution do not develop granules that withstand 
ramming in the mold. 

Examples of sands that granulated readily, gave conspicuously per- 
meable masses when molded, and were rather dense, or but slightly 
permeable when dry, are seen in samples 208, 240, 241, 242, 257, and 259. 
Comparing the mechanical analyses of these sands, Table XI and XII, 
it will be seen that they show a rather wide range in constitution. If 
the fines alone are considered, it will be seen that the 30-minute and 
15-minute silts and the sand passing the 200-mesh screen make 81, 27, 
22, 26, 33, and 25 per cent respectively of the several samples. Sand 
208, which shows an exceptionally large proportion of fines, is essentially 
a brass-molding sand, though not in general use, whereas the other samples 
referred to, ranging as low as 22 per cent of fines, are loams. 

In Table XVI samples of 240 to 251, inclusive, come from the same 
foundry, and represent both the molding sands and the core sands used. 



RESULTS OF PERMEABILITY TESTS 97 

Samples 245 to 251, inclusive, were cores made in the foundry by skilled 
molders, using their own core sand, their own binders, and baking the 
cores in the foundry ovens, so that these cores were comparable in every 
way with their regular practice. The remarkable uniformity in the per- 
meability of the cores made in triplicate illustrates the degree of skill 
attained by careful molders. 

Samples 240, 241, and 242, representing the molding sand in this 
found rv, are ordinary loams, as will be seen from the mechanical analyses, 
Table XII. They carry 2.5 to 4 per cent of clay, and the usual percentage 
of fine and coarse sizes, 'lie v. ork in this foundry was in malleable iron, 
mostly small castings, where the soundness of casting and smooth surface 
finish were important. 

This series of samples illustrates a close approach to the proper 
permeability for sands for light work in iron. It will be observed that 
the permeability in the green molds ranged from 3.4 to 6.4 per cent 
where hard rammed, and 4 to 10.2 per cent in air-dried molds. In 
Table XV it will be observed that this same series of sands, 240, 241, and 
242, showed good tensile strength both in green and in dry briquets. 

Samples 257, 258, and 259 represent another series of sands or loams 
in use for heavy work in grey iron. In this foundry a facing sand, 259, 
was used, and the molds were dried in all cases before casting. Number 
259 D is the most significant test of this series and shows a permeability 
of 5 per cent. The sand used in this foundry back of the facing sand 
259 was still more permeable and was used in the molds for heavy work. 
This sand, 259 D, as will be seen from Table XV, had a tensile strength, 
air-dried, of 17 pounds, and this was still further increased in actual 
foundry practice by spraying the inner surface of the mold, after com- 
pletion, with molasses water and drying the surface by the aid of a 
torch. In this manner sand 259 was made to stand up, and give a good 
skin, or surface finish, to castings weighing several tons. Samples 506 
to 511, inclusive, represent the type of core sand used in this foundry, 
in connection with the molding loams, 257, 258, and 259. 

Sample 260 illustrates the permeability required for brass sand for 
heavy work, viz., 2.8 per cent. This sand was used in railroad shops 
for casting brass journals. weighing 220 pounds for locomotives and gave 
exceptionally good results. Soundness of the casting was the essential 
feature in this work. It will be noted that the permeability is much less 
than for iron work. Samples 512, 513, and 514 are cores made from 
the core sand used in this railroad shop in connection with the brass 
sand, 261, the permeability of the cores being 25 to 27 per cent. 

Sample 292, which was used for ornamental bronze work, shows 
only 1.8 per cent permeability, but the castings here were for the most 



98 THE FOUNDRY SANDS OF MINNESOTA 

part light and a backing sand having a greater permeability was generally 
used. 

Samples 500 to 505 were cores made in a foundry doing grey iron 
work, mostly light castings. No information was furnished by the foun- 
dry as to the nature of the bond used, the proportion of the bond to the 
sand, or whether the six samples were all the same or represented differ- 
ent mixtures. The results of the test indicate that sample 503 was a 
different mixture from the others. 

Samples 506 to 511 represent a set of cores made by a foundry 
doing heavy founding in grey iron. Samples 507 and 508 were made, 
using a commercial binder with 1 part of binder to 30 parts of sand. 
Sample 509 used 1 part linseed oil to 45 parts of sand. Cores 510 and 
511 were probably triplicates of 509. 

Samples 515 to 523, inclusive, represent three series of triplicate 
cores, 515, 516, and 517 being one triplicate set; 518, 519, and 520 
being a second set; and 521, 522, and 523 being the third set. No in- 
formation was furnished by the foundry as to the kind of bond used, 
or the proportion of the bond to the sand, except that the cores repre- 
sented their regular foundry practice. The foundry was working in grey 
iron, mostly light casting. 

POROSITY TESTS 

In setting up Table XVI figures on porosity were omitted, but upon 
further consideration it seemed advisable to give them, and they are 
shown in Table XVII. 

The porosity percentages here given are of unequal value or accuracy 
by reason of the factors involved in some of the computations. With 
the samples of loose, dry sand the porosity was actually determined by the 
method described on page 41. For the neutral sands and loams the 
porosity percentages are dependable, but in the samples of floor sand or 
other used sand more or less foreign material was present such as chopped 
straw, sawdust, coal, etc., the specific gravity of which might be 1 or less 
than 1, or in any event less than the assumed specific gravity of 2.y. 
Again the used sands may or may not have contained particles of metal 
— brass, iron, etc., the specific gravity of which might be 5 or more ; thus 
introducing into the computations factors, the value of which was un- 
known. Different cylinders were used for weighing materials for per- 
meability and for porosity tests and the results of packing were not 
identical. 

The porosity of the molded cores was determined by taking the actual 
weight of the core, which was of the same volume as the cup V , Fig. 6, 
and computing the porosity after the method described on page 41. For 
cores that contained only natural sand or loam, and were air-dried, the 



POROSITY TESTS 



TABLE XVII 





Approximate 




Approximate 




Approximate 


Laboratory 


Porosity in 


Laboratory 


Porosity in 


Laboratory 


Porosity in 


Number of 


Percentage 


Number of 


Percentage 


Number of 


Percentage 


Sample 


by Volume 


Sample 


by Volume 


Sample 


by Volume 


215 stand- 




223-224 D 


37-8* 


258 C 


50.8 


ard sand 


33-5 


223-224 E 


43-7* 


258 D 




1 


35-3 


225 


38.6 


259 A 


45-7 


2 X 


40 


226 


37 


259 B 


46 


5 


2 Y 


39-4 


227 




259 c 


50 


9 


2 Z 


39* 


228 A 


35-4 


259 D 


46 


5* 


3 J 


47* 


228 B 


35-7 


260 


43 


7 


B 4 


42, 


228 C 


36.2 


292 


44 





c 4 


42 


228 D 


37-5 


293 


4i 


1 


9 A 


37 


228 E 


38.0 


297 






19 


33-6 


228 F 


39-8 


500 


37 


4 


21 


33-6* 


240 A 


41.8 


501 


38 





200 A 


49.4 


240 B 


46.3* 


502 


38 





200 B 


So. 4 


240 C 


47 


503 


44 


5 


200 C 


48.5 


240 D 


46 


505 


43 


5 


200 D 


48 


241 A 


39-7 


506 


4i 


3 


208 A 


47 


241 B 


42.9 


507 


42 


6 


208 B 


39 


241 D 


44 


508 


42 


6 


208 C 


44 


242 A 


42.6 


509 


37 


9 


215 


33-5 


242 B 


44.2* 


5io 


37 


9 


220 


35-4 


245 A 


37-8 


5ii 


37 


7 


221 A 


34 


246 B 


42.4 


512 


40 


8 


221 C 


35-9 


247 B 


42.6 


513 


4i 


1 


222 B 


34-9 


248 B 


43-4 


5i4 


40 


6 


222 c 


39* 


249 c 


43-4 


5i5 


37 





222, A 




250 C 


41-3 


516 


38 





22Z B 


33-6 


251 c 


40.8 


5i7 


37 


5 


224 A 


41.4 


257 A 


43-5 


5i8 


37 





224 B 




257 B 


47-3 


5i9 


37 


9 


224 C 


45-6* 


257 c 


49-5* 


520 


37 


7 


224 D 




257 D 


47-3* 


521 


40 


8 


223-224 A 


42.2* 


257 E 


44-7 


522 


39 


8 


223-224. B 




258 A 


43-7 


523 


40 


9 


222,-224 C 


35-4* 


258 B 


49.4 









* Indicates estimates or computations from incomplete data. 

porosity may be accepted as accurate, whereas for cores made of used 
sand the factor of variable specific gravity above mentioned may be in- 
volved and might affect the computation. 

Again in computing the porosity of the damp, "green" molded cores 
the water content was deducted from the weight of the core on the as- 
sumption that the water was absorbed by the clay and did not affect the 
porosity, which assumption might be open to question. 

It will be observed that the porosity of the standard sand used, viz., 
No. 215, is 33.5 per cent, and that practically all of the other samples 
run higher than that figure. The porosity of the air-dried cores also runs 
high, which is again consistent since they represent a structure in the 
mass abnormally open by reason of granulation. It is improbable that 
the variations in specific gravity above suggested appreciably affected the 
porosity computations, but this possibility must be taken into account. 



ioo THE FOUNDRY SANDS OF MINNESOTA 

As stated elsewhere, the percentage porosity is no measure, and not 
necessarily an index of the permeability ; in fact, in a series of sands 
ranging from very coarse to very fine, and otherwise the same in con- 
stitution, the permeability would necessarily be in reverse ratio to the 
percentage porosity, which is illustrated by the fact that standard sand 
33.5 per cent porous is 100 per cent permeable, whereas clay or heavy 
loam is 50 per cent or more porous, and practically impermeable. 

CONCLUSIONS 

From Table XVI it will be seen that there was a wide range in the 
permeability of the molding sands in actual practice. There is no very 
evident relation between the kind of founding done and the permeability 
of the mold. In foundries where the greatest care and most attention 
were given to the sand, the permeability of the mold ranged from 4.5 
to 7 per cent, and these limits may be taken as normal for iron foundings. 
In brass founding the permeability of the mold may be taken as 1.5 to 3 
per cent. 

The permeability of cores may be taken as ranging from 15 to 30 
per cent with little or no relation to the kind of work done, as evidenced 
from Table XVI. An exhaustive study of the problem would no doubt 
show a percentage permeability of the core varying with the nature of 
the founding done. 

The material preferred for core work is a quartz sand with the 
least possible amount of silts, and with little or no clay, since organic 
bonds that will burn out in the founding process, leaving the sand loose 
or incoherent, are essential in most work. The St. Peter sandstone affords 
an abundance of this type of sand well suited to ordinary founding. 
Examples of this sand are samples 1, 5, 21, 28, 31, 64, 90, 145, 262, and 
299, all in Table IX. 

Coarser quartz sand, sometimes desired for cores in heavy work in 
grey iron and for founding in steel, may be readily obtained from the 
Jordan formation, examples of which are samples 8, 19, 159, 166, 167, 
170, 173, and 175, Table X. 

The permeability tests of the molding sands, some of which are given 
in Table XVI, brought out many interesting facts in connection with the 
permeability of the same materials under different treatments, and the 
relative permeability of different materials under like treatment. For ex- 
ample, with the same loam or sand dampened and molded, some loams 
showed greater permeability in the green or damp condition than in the 
air-dried condition, whereas with other loams the reverse was true. Again 
some sands or loamy sands showed greater permeability in the loose dry 



CONCLUSIONS 101 

state than in the damp molded state, whereas other sand showed the re- 
verse. Such seemingly erratic behavior was evidently due to peculiarities 
in the physical constitution of the particular materials by virtue of which 
some readily took on a granular structure, whereas others did not, under 
the same treatment. 

Tests of the cores made in different foundries indicated that the per- 
gonal equation of the molcler was an apparent factor in the permeability. 
Several cores made by the same molder, from the same sand at the same 
time under supposedly uniform conditions showed wide variations in the 
permeability of individual cores, whereas other groups of cores made by 
other molders in other foundries from similar materials showed remark- 
ably uniform permeability. It is of course possible that some peculiarity 
in the foundry practice might account for these differences in part, and 
the personal equation of the molder may not have been wholly responsible. 
However, on the whole, the skill of the molder seemed in some instances 
to be the chief factor. 

In sands where no artificial bond was used, and dependence was 
placed on the clay or other material inherent in the natural sand, varia- 
tions were observed in the permeability which were not directly related 
to the percentage of clay present ; indicating that the nature of the clay 
itself was a factor in permeability. 

The percentage of silts in the sand affected the permeability of the 
sample differently in different types of sand, and also gave different 
results with the same sand in the various foundries. In the latter case 
the results would seem to be due to the foundry practice in handling the 
sand ; but in the former, the difference in the behavior of the silts ap- 
peared to be due to some peculiarity inherent in the sand or the silts. 
Some sands, if properly handled, get rid of the silts by granulation, in 
which case the silts adhere to the larger grain, or the silts themselves 
gather into granules, giving an open permeable sand ; whereas in other 
sands, no matter how handled, the silts remain as independent silty ma- 
terial that clogs the interstitial spaces, decreasing permeability. 



INDEX 



"A" Analyses 36 

Acknowledgments 3 

Air engine 34 

Albany sand, cost of 1 

Algonkian in Minnesota 13 

Alkalies, effect on sand 7 

Allen, V. T., acknowledgments to 3 

Alloys 28 

American Brake Shoe Company 32 

American Gas Machine Company 31 

American Hoist and Derrick Foundry 32 

Analyses of clays 78 

Analyses of loess 30 

Apparatus for permeability tests 44 

Archean in Minnesota 13 

Area covered in report 1 

Armstrong, H. K., acknowledgments to... 3 

Arrangement of pores in sands 48 

Artificial or mixed sands 8 

Austin Foundry 31 

Average mesh 38 

"B" analyses 36 

Barium chloride treatment of clay 35 

Barnes Bluff, loess of 29 

Bellechester, clay of 23 

Bonding power tests 39 

Brass molding sand 28 

Brass sands, tests of 66, 68, 69 

Briquets for strength tests 40 

Brown's Valley, clay of 25 

Cambrian, sands of 14 

Cla y 35, 75 

Clay Bank, loess of 29 

Clay bond 34 

Clay content of loams 72 

Clay in sands, function of 34 

Clays, mechanical analyses of 32 

Cloquet Foundry 31 

Clyde Iron Works 31 

Coal screenings used in sands 8 

Colloids in clay j6 

Commutator Company 32 

Comparison of analyses ^7 

Comparative diameter of grains 38 

Comparative mesh 38 

Conclusions 100 

Constant air pressure apparatus 44 

Core sand, tests for 60 

Coteau des Prairies 10 

Courtland, quartzite at 18 

Cretaceous clay 23 

Cretaceous in Minnesota 14 

Cretaceous sands, tests of 58 

Cretaceous sandstone 22 

Crookston Iron Works 31 



Crown Iron Works 32 

Cylinders for permeability tests 43, 47 

Cylinders for testing sands 42 

Decorah shale 22 

Deflocculation of clay 32 

Deterioration of sand with heat 84 

Devonian in Minnesota 14 

Doctoring sands 8, 70 

Dresbach, sands of 14 

Drift, clay in 25 

tesls of sands of 68 

Dry methods of sand analysis 34, 35 

Duluth Brass Works 31 

Duluth Foundry and Faucet Company.... 31 

Duluth Iron Works 31 

Duluth, recent sands used at 30 

sands used at 13, 30 

Eagle Foundry Company 32 

Effective size of grain 52 

Emmons, W. H., acknowledgments to 3 

Engine, hot air 34 

Erikson, H. A., acknowledgments to 3 

Ess Bros 31 

Faribault Machine Shop 31 

Fairmont Gas Engine Company 31 

Fatigue of bond 85 

Fergus Falls Iron Works 31 

Fifteen-minute silts 35 

Flask 5 

Flour City Ornamental Iron Company. ... 32 

Formations in Minnesota 12 

Fort Snelling, analysis of sands of 22 

Founding, history of 3 

Foundry sands, tests of 32 

Frame 5 

Franconia, Minnesota 14 

French sand, test of 69 

French sands 1 

Fusion tests 57 

Gas Traction Foundry 32 

Gate City Iron Works 32 

Geologic column of Minnesota 12 

Geology of Minnesota 9 

Gillett-Eaton and Squire Company 31 

Glacial deposits 25 

Glacial loams 26 

Glucose, use of in sand 8 

Glue used with sand 8 

Gortner, R. A., acknowledgments to 3 

Granite City Iron Works 31 

Granulation 37, 52 

Gravel plains 10 



104 



INDEX 



Grout, F. F., acknowledgments to 3 

analyses by. . . . 30 

Herzog Foundry 32 

Hinckley sandstone 13 

Hot air engine 34 

Houston County, sandstone in 20 

Hunter, W. H., acknowledgments to 3 

Huronian in Minnesota 13 

Illinois, brass sand from 28 

Imported sand, tests of 69 

Jasper, quarries at 18 

Jordan formation, sands of 14, 19, 20 

tests 62, 64, 65 

Kansas drift in Minnesota 16 

Kasota, sand pits at 20 

sandstone 20 

Kentucky, brass sands from 28 

Kettle River formation 18, 19 

sands, tests for 65 

sandstone 13 

King, F. H., work of. 41 

Knapp, Mrs. W. C, acknowledgments to. . 3 

Laboratory tests 32 

Lake Agassiz n 

Lake beaches 28 

Little Giant Company 31 

Loams 26, 27 

mechanical analyses 73 

tests of 69, 70, 73, 74 

Loam soils 27 

reclaiming land after removing loam. . 27 

Loess 28, 39 

analyses of 30 

map of 28 

tests of 68,69 

Luverne, quartzite at 18 

Machine for testing permeability of sands 44 

Mankato clay 24 

Mankato Manufacturing Company 31 

Mechanical analyses 58 

Merriam Junction sand pits 20 

Mineral analyses 58 

Minneapolis, sandstone at 21 

Minneapolis Steel and Machinery Company 32 

Minnesota loess deposits, map of 28 

Minnesota Radiator Company 31 

Missouri, brass sand from 28 

Mixed sands 8 

Moisture content, tests of 57 

Molasses, use of, with sand 8 

Molding sands 6 

Molds 5 

Moraines 10 

clays of 25 

Mounting sands for testing permeability. .46, 47 



National Iron Company 31 

New Jersey, loam sands reclaimed 27 

New Owatonna Manufacturing Company.. 31 

New Prague Foundry Company 31 

New Ulm, clay of 25 

quartzite at 18 

New Winona Manufacturing Company. ... 32 

North Star Iron Works 31 

Northern Malleable Iron Works 32 

Northfield, sandstone of 21 

Nutting Truck Company 31 

Oliver Mining Company 31 

Oneota sands, tests of 61 

Openings in sands 50, 56 

Ordovician in Minnesota 14 

Ortonville Foundry 31 

Ottawa, Illinois, sands 20 

Ottawa, Minnesota, sands 20 

Outline of geology of Minnesota 9 

Outwash plains 10 

deposits of 26 

loams of 26, 27 

Parker and Topping Company 31 

Pattern -making 4 

Peavey, C. W 31 

Percentage of permeability 54 

Permeability 43 

relation to structure 48 

test of 89 

Pipestone County, Kansas drift of 16 

quartzite 18 

Plasticity of clays 76 

Pleistocene in Minnesota 16 

Pore space 49 

relation of granulation to 52 

Porosity 48 

tests of 41,99 

Quartzite screenings 81 

Railroad administration acts concerning 

sand 1 

Recent deposits 30 

Recent formations in Minnesota 17 

Reclaiming loams 27 

Red Clastic series 13 

Red Wing Iron Works 31 

Red Wing, Minnesota, clay of 23, 24 

loess of 29 

sandstone of 21 

Relation of meshes in screens and openings 54 

Riser, defined 5 

Rock County, drift in 16 

Rock flour 26 

St. Cloud Iron Works 31 

St. Lawrence formation 14 

St. Paul Foundry Company 32 

St. Paul, sandstone of 21 

St. Peter sandstone 21 

St. Peter sands, beds of 60 



INDEX 



105 



Sands, analyses of 32 

Sandstone, quarries at 18 

Sawdust used with sand. . 8 

Screen analyses 33 

Sea coal 8 

Searching of sand by liquid metals 8 

Shakopee sand, tests of 61 

Shaper, W. A., acknowledgments to 3 

Silt 26, 35 

Sioux Falls, quartzite at 18 

Sioux quartzite 17, 19 

Size of clay particles 37 

Size of silt particles 37 

Sized sands, permeability of 95 

Slichter's experiments 48 

Smith Grubber Company 31 

Soo Line Railway shops 32 

South Bend, clay near 25 

Southeast Minnesota 11 

Steel sands, tests for 62 

Strength of clay yy 

Structure, relation of porosity to 48 

Synthetic sands 72 



Temperature of molding 7 

Tensile strength, tests of 39, yy, 82, 86 

Testing apparatus 39 

Tests 2,8 

for porosity 41 

for strength 37, yy, 82, 86 

Texture, relation of porosity to 88 

Thief River Falls Iron Works 32 

Thirty-minute silt 35 

Tubal-Cain 4 

Union Brass Works 7,2 

University of Minnesota Foundry 32 

Valley Iron Works 32 

Virginia Foundry Company 32 

Voids in sands 50 

Wet methods of sand analysis 34,35,36 

Winona Machinery and Foundry Company 32 

Winter and Company 31 

Wire screens 33 

Wisconsin drift 17 

Zumbro River, rocks of 19